Radiologic Technologists and Technicians

29-2034.00
Median wage $80,110/yr230,490 employed (US)Rank #632 of 923 scored · top 68% by substitution

Take x-rays and CAT scans or administer nonradioactive materials into patient's bloodstream for diagnostic or research purposes. Includes radiologic technologists and technicians who specialize in other scanning modalities.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution22
Exposure21
Augmentation48

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

30 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

3%

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%21

panel mean rating 1.8/5 → substitution pressure 21/100

Technical feasibility todayw 20%21

panel mean rating 1.8/5 → substitution pressure 21/100

Cost vs. human wagew 15%20

panel mean rating 1.8/5 → substitution pressure 20/100

Adoption barriersw 20%inverted — strong barriers lower the score23

panel mean rating 4.1/5 (barrier strength) → substitution pressure 23/100

Sector adoption velocityw 10%23

panel mean rating 1.9/5 → substitution pressure 23/100

Task breakdown (30 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Process exposed radiographs using film processors or computer generated methods.

81

CI 75–86 · exposure 84 · augmentation 63 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5Digital and automated radiographic processing has achieved near-universal adoption in medical imaging facilities across developed healthcare systems, with PACS integration now standard practice rather than innovation.
Sector adoption velocityclaude-sonnet-54/5Healthcare imaging has broadly shifted to digital radiography and PACS over the past two decades, representing fast, deep adoption of this specific processing step even though other radiology tasks lag.
Augmentation potentialclaude-haiku-4-5-202510014/5Automated processing and PACS dashboards substantially augment technician productivity by eliminating manual film handling, enabling rapid retrieval and manipulation of images, and flagging quality issues for review, while technicians retain quality control and exception handling roles.
Augmentation potentialclaude-sonnet-53/5Automated processing assists technologists by speeding workflow and reducing manual steps, but it is a narrow supporting function rather than a transformative augmentation of the technologist's broader role.
Task automatabilityclaude-haiku-4-5-202510014/5Modern PACS (Picture Archiving and Communication Systems) and automated film processors can handle the majority of post-exposure processing—standardizing, archiving, and initial quality checks—with significant time savings. However, quality assessment and handling of failed exposures still requires human intervention, preventing a full 5 rating.
Task automatabilityclaude-sonnet-54/5Film/image processing is largely mechanical and computerized (CR/DR systems auto-process images), with minimal human judgment needed once exposure is captured, so most of this specific step can be automated with existing digital radiography workflows.
Adoption barriersclaude-haiku-4-5-202510013/5Regulatory standards (DICOM, HL7) and quality assurance protocols require oversight and validation, and many facilities maintain human-in-the-loop QA processes. However, no legal requirement mandates a licensed human perform automated processing itself.
Adoption barriersclaude-sonnet-52/5No licensure requirement specifically for image processing itself, though it is embedded within a licensed technologist's overall workflow and equipment must meet regulatory/quality standards.
Cost vs. human wageclaude-haiku-4-5-202510015/5Automated processing systems (both film and digital) have minimal per-image marginal cost once deployed, yielding orders-of-magnitude savings compared to the loaded wage of a technician performing manual processing on each radiograph.
Cost vs. human wageclaude-sonnet-54/5Automated digital processing systems are far cheaper per image than manual darkroom film processing, though hardware/software integration and maintenance costs keep it from being an extreme 10x+ differential in all settings.
Technical feasibility todayclaude-haiku-4-5-202510015/5Automated film processors and PACS systems are mature, widely deployed products in hospitals and imaging centers globally, reliably performing digitization, archival, and standardized processing at scale in production environments.
Technical feasibility todayclaude-sonnet-55/5Digital radiography and PACS systems already automate image processing reliably in production across most modern hospitals and clinics, replacing manual film development almost entirely.

Record, process, and maintain patient data or treatment records and prepare reports.

64

CI 50–79 · exposure 62 · augmentation 75 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Healthcare IT adoption of AI-driven record systems, NLP report processing, and EHR automation is accelerating rapidly in hospital networks and imaging centers, with measurable displacement of manual data entry and clerical work already underway.
Sector adoption velocityclaude-sonnet-53/5Healthcare documentation tools are increasingly adopted, but healthcare overall lags behind fully digital sectors like finance or software in AI integration depth.
Augmentation potentialclaude-haiku-4-5-202510014/5AI assistants enhance technologists' productivity by auto-populating fields, flagging missing data, generating draft reports for review, and organizing records, allowing them to focus on higher-value clinical and procedural tasks while remaining in the loop.
Augmentation potentialclaude-sonnet-54/5AI-assisted transcription, template generation, and data extraction meaningfully speed up documentation tasks while technologists retain oversight of accuracy.
Task automatabilityclaude-haiku-4-5-202510014/5AI can reliably extract structured data from imaging reports, populate EHR fields, generate standardized report sections, and organize patient metadata with high accuracy. However, some medical decision documentation and quality assurance may require human oversight, preventing a full 5 rating.
Task automatabilityclaude-sonnet-53/5AI can transcribe, structure, and populate documentation from dictation or structured inputs, but final data entry into clinical records still typically requires human verification for accuracy and compliance.
Adoption barriersclaude-haiku-4-5-202510012/5While HIPAA and data governance require oversight, there is no legal requirement for a licensed radiologic technologist to personally record or format data; oversight can be lightweight. No hard regulatory barrier prevents substitution of this documentation work.
Adoption barriersclaude-sonnet-53/5Patient record accuracy, HIPAA compliance, and clinical accountability create moderate barriers, though data entry itself isn't inherently a licensed-only act.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-driven record processing, NLP extraction, and report generation cost a fraction of full-time technician labor for the same volume and turnaround, achieving order-of-magnitude cost savings per task-equivalent when amortized across patient volume.
Cost vs. human wageclaude-sonnet-53/5Documentation software reduces time spent but still requires licensed staff oversight and integration costs, keeping savings moderate rather than order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed systems (EHR integrations, NLP-based report generation, medical record automation) perform these tasks reliably in production at scale in many hospitals and imaging centers. Mature products exist, though integration complexity and institutional variation create minor implementation friction.
Technical feasibility todayclaude-sonnet-53/5EHR-integrated speech recognition and documentation assistants are deployed in many radiology departments, but full end-to-end automation of record maintenance with no human review is not standard.

Perform general administrative tasks, such as answering phones, scheduling patient appointments, or pulling and filing films.

64

CI 54–75 · exposure 62 · augmentation 63 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Healthcare organizations, particularly larger systems and hospital networks, are actively adopting AI-driven scheduling and phone automation. Adoption is measurable in production environments, though smaller practices lag.
Sector adoption velocityclaude-sonnet-53/5Healthcare administrative functions are adopting scheduling/chat automation at a moderate pace, generally slower than pure information-sector adoption due to compliance and legacy IT systems.
Augmentation potentialclaude-haiku-4-5-202510013/5AI assists radiologic technologists by handling routine calls and appointment slots, freeing them for clinical duties, but does not transform the full administrative workflow given physical file handling requirements.
Augmentation potentialclaude-sonnet-54/5AI scheduling assistants, digital filing systems, and call-routing tools meaningfully reduce administrative burden on technologists, letting them focus more on clinical duties.
Task automatabilityclaude-haiku-4-5-202510013/5Phone answering and basic appointment scheduling can be automated by AI systems (chatbots, voice assistants), but pulling and filing physical films cannot. Hybrid automation of ~50% of the workload with significant setup is feasible, meeting the automatability threshold.
Task automatabilityclaude-sonnet-54/5Scheduling, phone triage, and digital file/records management are well within current AI capabilities (chatbots, scheduling software, EHR integrations) that can save significant time on these administrative subtasks.
Adoption barriersclaude-haiku-4-5-202510013/5Some friction exists from HIPAA compliance requirements, patient preference for human contact, and integration with legacy scheduling systems. However, no licensing requirement mandates a human perform these administrative tasks, limiting hard barriers.
Adoption barriersclaude-sonnet-52/5These are non-clinical, non-licensed tasks with minimal regulatory or liability barriers, though some patient-preference friction and integration with legacy systems exists.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI-driven scheduling and phone systems have low marginal inference costs compared to human receptionist wages, though integration and healthcare-specific customization add overhead. Overall, AI is substantially cheaper per task equivalent.
Cost vs. human wageclaude-sonnet-54/5Automated scheduling and phone-answering systems are substantially cheaper than paying technologist time or dedicated administrative staff for these routine tasks.
Technical feasibility todayclaude-haiku-4-5-202510013/5Chatbots for answering phones and scheduling exist in production healthcare settings, but often with error rates requiring human override. Physical file management remains manual. Products are deployed but with material limitations in scope and reliability.
Technical feasibility todayclaude-sonnet-54/5Deployed scheduling assistants, automated phone systems, and digital PACS-based film management are already in production use across many healthcare systems, though physical film pulling may still occur in some settings.

Maintain a current file of examination protocols.

47

CI 29–65 · exposure 45 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Radiology departments are moderately digitized but protocol maintenance remains largely manual and institutional in many settings. While some large health systems use centralized protocol databases, widespread adoption of AI-driven protocol management is still nascent and not yet standard practice.
Sector adoption velocityclaude-sonnet-52/5Healthcare/imaging departments are moderate-to-slow adopters of AI for administrative/documentation tasks, with most AI investment focused on image analysis rather than protocol management workflows.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can substantially assist by automatically indexing protocols, flagging outdated versions, tracking guideline updates, and organizing documents by body part or modality. These augmentations would significantly speed up a technician's ability to locate and maintain current protocols while the human retains final oversight of accuracy and compliance.
Augmentation potentialclaude-sonnet-54/5AI tools can effectively help technologists draft, format, cross-reference, and update protocol documents, substantially easing the administrative burden while the technologist retains final oversight.
Task automatabilityclaude-haiku-4-5-202510012/5Creating and organizing examination protocols requires domain expertise and understanding of clinical guidelines, but maintaining an existing file of protocols—tracking versions, updating references, organizing documents—involves routine document management that AI could partially automate. However, clinical protocol validation and accuracy checking would still require human oversight, limiting time savings to well below 50%.
Task automatabilityclaude-sonnet-54/5Maintaining and updating a protocol file is largely a documentation/knowledge management task—AI can draft, update, and organize protocol documents from source guidelines with human review, saving significant time.'
Adoption barriersclaude-haiku-4-5-202510014/5Medical protocols are regulated under various standards (ACR, ASRM, facility accreditation) and any changes must meet compliance requirements and institutional oversight policies. Radiologic technologists and radiologists must verify protocol accuracy for patient safety and liability reasons, creating a strong organizational and regulatory barrier to full automation.
Adoption barriersclaude-sonnet-52/5No strict licensing requirement for maintaining a protocol file, though clinical protocols often require sign-off from a radiologist or lead technologist for accuracy and safety compliance.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI-driven document management and version control systems are inexpensive to deploy compared to a technician's loaded wage, especially for routine file maintenance, indexing, and change-tracking tasks. The infrastructure cost is typically a small fraction of annual technician compensation.
Cost vs. human wageclaude-sonnet-53/5AI-assisted drafting and updating could be cheap per instance, but integration with clinical systems, verification against current standards, and low task frequency mean overall cost savings are moderate rather than dramatic.
Technical feasibility todayclaude-haiku-4-5-202510013/5Document management and version-control systems exist and can track protocol changes automatically, but deployed radiology-specific protocol management solutions have mixed adoption and often require significant human curation. General AI document tools can assist but lack specialized medical protocol validation.
Technical feasibility todayclaude-sonnet-52/5While generic document management and LLM drafting tools exist, no widely deployed radiology-specific product autonomously maintains protocol files in production settings today; this is typically done manually via department wikis or shared drives.

Review and evaluate developed x-rays, video tape, or computer-generated information to determine if images are satisfactory for diagnostic purposes.

37

CI 25–49 · exposure 38 · augmentation 63 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Hospitals and imaging centers are piloting AI quality-control tools, but adoption is uneven and mostly adjunctive rather than displacement; many institutions still rely on manual review, and integration friction slows deployment in legacy PACS environments.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging is a heavily regulated, moderate-digitization sector where AI QA tools are in early pilot stages rather than widespread production use for this specific evaluative step.
Augmentation potentialclaude-haiku-4-5-202510014/5AI substantially augments technicians by automating routine flagging of obvious defects, summarizing image metadata, and surfacing outliers for focused human review, thereby raising throughput and reducing fatigue-driven errors while preserving human oversight.
Augmentation potentialclaude-sonnet-53/5AI-based image quality flags and automated exposure indicators can help technologists catch problems faster, offering useful but partial assistance to this judgment task.
Task automatabilityclaude-haiku-4-5-202510013/5AI can detect technical defects (positioning, exposure, blur) and flag unsatisfactory images with reasonable accuracy, automating perhaps 40–60% of the quality-control workflow. However, radiologists retain final authority on diagnostic suitability, and edge cases require human judgment, preventing full end-to-end automation at equal quality.
Task automatabilityclaude-sonnet-52/5Image quality assessment for diagnostic adequacy requires physical positioning knowledge, artifact recognition, and contextual judgment tied to the specific patient and exam; AI can flag some technical issues but cannot fully replace the technologist's real-time decision loop with the patient present.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: radiologists and credentialed technicians retain clinical and legal responsibility for image adequacy; liability risk and regulatory oversight (FDA, HIPAA, accreditation) mean that unsupervised substitution is not legally permissible. Human sign-off is mandatory.
Adoption barriersclaude-sonnet-54/5Diagnostic imaging quality control is tied to licensed radiologic technologist responsibilities and regulatory/accreditation requirements, and patient presence necessitates a human in the room to redo images if needed.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI inference for image screening is cheap (pennies per image after amortized setup); integrating with PACS and oversight adds modest cost, but still substantially undercuts technician labor ($40–60/hour loaded) for systematic screening and triage.
Cost vs. human wageclaude-sonnet-52/5Any AI QA tool would be an add-on to existing workflow and equipment, requiring integration and oversight costs, while the technologist is already on-site performing the exam, so cost savings are limited.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI-based image quality assessment tools exist in research and limited clinical deployment (e.g., vendor-integrated solutions), but error rates on complex or borderline cases remain material. Deployed systems typically flag outliers rather than replace human review entirely, indicating narrow scope relative to production maturity.
Technical feasibility todayclaude-sonnet-52/5Some QA software exists that flags exposure or motion artifacts, but no deployed product independently determines full diagnostic adequacy across modalities in place of the technologist at the point of care.

Operate digital picture archiving communications systems.

34

CI 23–46 · exposure 38 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare has been slow to adopt autonomous automation of PACS workflows compared to other sectors; adoption is heavily constrained by regulatory approval pathways, resistance to perceived reduced human oversight in critical clinical processes, and entrenched vendor PACS ecosystems.
Sector adoption velocityclaude-sonnet-53/5Healthcare imaging IT has moderate AI adoption with pilots for triage and prioritization tools becoming more common, but full operational control of PACS by AI remains rare in production settings.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist technicians through intelligent pre-filing suggestions, automated quality checks flagging corrupted images, and predictive routing recommendations that improve operator efficiency, though the human remains essential for final validation and exception handling.
Augmentation potentialclaude-sonnet-54/5AI-based worklist prioritization, image quality checks, and automated tagging meaningfully speed up technologist workflow within PACS while the human remains responsible for system operation.
Task automatabilityclaude-haiku-4-5-202510012/5Operating PACS (digital picture archiving systems) involves routine data entry, image routing, and system navigation. While some UI interactions could be automated, the task requires contextual judgment about proper patient/study linkage, error correction, and system troubleshooting that current AI struggles with reliably in a healthcare setting.
Task automatabilityclaude-sonnet-53/5AI can automate portions of PACS workflow like image routing, prioritization, and worklist management, but full operation including troubleshooting, quality control, and system navigation still requires human interaction with hardware/software interfaces.
Adoption barriersclaude-haiku-4-5-202510014/5HIPAA compliance, medical device regulation (FDA oversight of PACS as Class II/III systems), and institutional liability requirements mean that automated image handling and patient data routing typically must be validated and signed off by licensed personnel, creating significant legal and regulatory barriers.
Adoption barriersclaude-sonnet-53/5Handling patient imaging data involves HIPAA compliance, hospital IT governance, and credentialing requirements for accessing and managing medical records, creating moderate regulatory and organizational friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI infrastructure for PACS operation (licensing, integration, validation, ongoing maintenance) plus required human oversight approaches or exceeds the loaded cost of a technician for routine PACS tasks, especially when accounting for liability and error-correction overhead.
Cost vs. human wageclaude-sonnet-52/5PACS integration with AI tools requires significant IT infrastructure, licensing, and maintenance costs that are not dramatically cheaper than the marginal labor cost of a technologist operating existing systems.
Technical feasibility todayclaude-haiku-4-5-202510012/5Existing RPA and AI products can handle template-based PACS operations (login, image retrieval, basic routing), but production deployments remain limited due to the need for clinical validation and the critical nature of image integrity. Most healthcare PACS workflows still require human oversight.
Technical feasibility todayclaude-sonnet-53/5PACS systems already incorporate AI-driven triage and worklist prioritization in production radiology departments, but full autonomous operation of the archiving/communications system itself is not deployed without a technologist present.

Perform supervisory duties, such as developing departmental operating budget, coordinating purchases of supplies or equipment, or preparing work schedules.

28

CI 25–30 · exposure 25 · augmentation 63 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare organizations lag in adopting autonomous AI for supervisory functions due to compliance sensitivity and cultural conservatism; most adoption remains limited to scheduling assistance tools with heavy human review rather than autonomous decision-making.
Sector adoption velocityclaude-sonnet-52/5Healthcare administrative functions are adopting AI slower than finance or tech; departmental management tasks in imaging centers remain largely manual with pilot-level tool use.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist with budget forecasting, supply cost analysis, and schedule optimization recommendations, but the supervisor must ultimately validate and approve changes based on departmental context and compliance requirements.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist with drafting budgets, analyzing supply costs, and generating draft schedules, giving supervisors a significant productivity boost while retaining decision authority.
Task automatabilityclaude-haiku-4-5-202510012/5Budget development and work scheduling involve routine data entry and rule-based logic that AI can partially assist with, but departmental oversight requires domain judgment, stakeholder negotiation, and contextual decision-making that current AI systems cannot handle end-to-end reliably.
Task automatabilityclaude-sonnet-52/5Budgeting, scheduling, and procurement coordination involve judgment, negotiation, and organizational context that current AI can support but not fully execute end-to-end reliably.'
Adoption barriersclaude-haiku-4-5-202510014/5Radiology departments operate under regulatory and accreditation requirements (Joint Commission, CMS) that typically mandate human supervisory accountability for budget decisions, staffing, and compliance; liability and legal responsibility create hard barriers to full automation.
Adoption barriersclaude-sonnet-53/5No licensing requirement forces a human to do budgeting or scheduling, but organizational accountability, HR policies, and managerial authority create real friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI tools for budget and scheduling assistance still require significant human oversight and integration costs, making the total cost comparable to or potentially higher than having a human supervisor perform the work directly.
Cost vs. human wageclaude-sonnet-52/5AI tools can cut some drafting/analysis time cheaply, but human oversight, approvals, and contextual decisions still dominate cost, keeping savings modest relative to a supervisor's wage.
Technical feasibility todayclaude-haiku-4-5-202510012/5No mature product performs full supervisory budget and scheduling tasks in production for radiology departments; fragments exist (budget templates, scheduling software) but they require substantial human configuration and approval rather than autonomous execution.
Technical feasibility todayclaude-sonnet-52/5Generic scheduling and budgeting software (some AI-enhanced) exists, but no deployed product autonomously manages a radiology department's supervisory administrative duties in production.

Take thorough and accurate patient medical histories.

28

CI 25–30 · exposure 25 · augmentation 63 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of fully autonomous AI for clinical history-taking remains limited and mostly in pilot phases; most radiology departments still use human technicians and human-reviewed EHR systems rather than AI-driven history collection, reflecting organizational conservatism and regulatory caution in clinical settings.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially imaging departments, adopts digital intake and NLP tools slowly due to compliance, EHR integration complexity, and safety-critical nature of the data collected.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist with structured prompts, question suggestions, and documentation templates that speed up note-taking and ensure completeness, but the human technician remains in the loop conducting and validating the interview; this represents useful but partial productivity gain.
Augmentation potentialclaude-sonnet-54/5AI-assisted intake forms, voice transcription, and EHR auto-population can meaningfully speed up documentation and reduce technologist workload while they remain responsible for verification and judgment.
Task automatabilityclaude-haiku-4-5-202510012/5Taking thorough patient histories requires nuanced interviewing, contextual understanding, and the ability to probe for relevant symptoms based on patient responses. While AI can assist with structured data entry and prompt suggestions, it cannot reliably conduct the full conversational interview without human judgment and medical reasoning, falling well short of the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5AI can help structure or transcribe history-taking, but eliciting nuanced, accurate patient history involves rapport, follow-up questioning, and physical presence that current systems cannot fully replace end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Medical history-taking carries regulatory and liability weight: inaccurate histories can compromise diagnosis and treatment, clinical judgment about symptom significance is expected from trained staff, and many healthcare settings and insurers require a human clinician or certified technician to sign off on patient intake for liability and compliance reasons.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement mandates a human collect history, but clinical liability, need for accurate contraindication screening (e.g., contrast allergies, implants), and patient trust create meaningful friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI systems (chatbots, structured intake forms) still require significant human oversight and correction, and the cost of integration, maintenance, and liability review often exceeds the savings from partial automation relative to a technician's loaded wage.
Cost vs. human wageclaude-sonnet-52/5Digital intake tools are cheap, but they still require human verification and follow-up, so total cost savings versus a trained technologist doing this directly are modest.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably takes full medical histories from patients end-to-end in production settings. Clinical documentation assistants exist for note-taking, but they require a human to conduct the interview and make clinical decisions about what is relevant; no mature system performs the entire history-taking task autonomously.
Technical feasibility todayclaude-sonnet-52/5Some deployed tools (voice-to-text intake forms, chatbot pre-visit questionnaires) exist but are narrow in scope and not a substitute for a technologist's judgment-based history taking during the actual imaging encounter.

Determine patients' x-ray needs by reading requests or instructions from physicians.

25

CI 25–25 · exposure 25 · augmentation 50 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare imaging facilities are moderately digitized and use RIS/PACS systems, but adoption of autonomous order-determination AI remains limited; most pilots focus on image analysis (reading), not intake/request validation. Sector-wide deployment of AI for order verification is still nascent.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging departments adopt AI slowly for clinical workflow decisions due to regulatory, liability, and interoperability constraints, with pilots more common than production deployment.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist radiologic technologists by automatically extracting and flagging relevant data from physician requests, prompting protocol confirmation, and checking for missing information, meaningfully reducing manual data entry and verification time while the technologist retains final decision authority.
Augmentation potentialclaude-sonnet-53/5AI-based decision support and NLP tools can help parse physician requests and suggest protocols, improving efficiency while the technologist retains final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5AI can parse structured physician requests and extract imaging orders with high accuracy, but determining appropriateness requires clinical judgment about patient history, contraindications, and protocol selection—tasks still requiring human oversight. Partial automation of form parsing is achievable, but end-to-end autonomous determination of imaging needs falls short of the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5Interpreting physician orders and translating them into appropriate imaging protocols requires clinical judgment, patient-specific context, and coordination that current AI cannot fully replace end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Radiologic technologists are licensed professionals whose scope includes independent verification of physician orders and assessment of imaging appropriateness; legal and regulatory frameworks (state licensure, facility credentialing) require licensed personnel to validate orders before imaging proceeds. Liability and patient safety concerns create strong friction against full automation.
Adoption barriersclaude-sonnet-54/5Radiologic technologists are licensed professionals, and determining correct imaging protocols involves patient safety and regulatory oversight (radiation dose, exam appropriateness) requiring human accountability.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-based form parsing and order-management systems carry integration and oversight costs, but radiologic technologists' verification time remains necessary. The cost savings from automating document reading alone do not offset the full technologist wage for this task.
Cost vs. human wageclaude-sonnet-52/5AI-assisted order interpretation tools require integration, oversight, and error-checking by a licensed technologist, so cost savings versus a trained human are modest at best.
Technical feasibility todayclaude-haiku-4-5-202510012/5Document parsing and order extraction tools exist and work reliably on well-structured forms, but clinical systems still require radiologic technologists to verify orders, check for gaps, and confirm appropriateness against patient context. No deployed product autonomously determines x-ray needs without human review.
Technical feasibility todayclaude-sonnet-52/5Some clinical decision-support and order-entry systems can flag protocol suggestions, but no deployed product reliably determines full x-ray needs from physician requests without technologist review.

Key commands and data into computer to document and specify scan sequences, adjust transmitters and receivers, or photograph certain images.

25

CI 25–25 · exposure 25 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While PACS and RIS systems have digitized workflow, actual adoption of autonomous equipment adjustment and protocol selection remains minimal in practice. Healthcare sector moves conservatively on clinical task automation, with most implementations limited to administrative support.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging is a highly regulated, physically-mediated sector with slower AI adoption for hands-on operational tasks, though AI is used more in image interpretation than in scan operation itself.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist with recommending standard protocols based on order information, flagging quality issues in captured images, and automating routine documentation, meaningfully supporting technician productivity while they retain control over equipment operation and clinical decisions.
Augmentation potentialclaude-sonnet-53/5Voice recognition and templated data entry systems can speed up documentation and command entry, offering moderate productivity gains while the technologist remains fully in control of the scan process.
Task automatabilityclaude-haiku-4-5-202510012/5While data entry and image capture are partially automatable, the task requires technical judgment about scan sequences and receiver/transmitter adjustments based on patient anatomy and clinical context. Current AI can assist with documentation but cannot independently make the clinical protocol decisions that define this role.
Task automatabilityclaude-sonnet-52/5Data entry portions could be partially automated, but the physical adjustment of transmitters/receivers and hands-on scan setup requires human presence with equipment and patients, limiting end-to-end automation.'
Adoption barriersclaude-haiku-4-5-202510014/5Radiologic technologists must be licensed/certified in most jurisdictions, and actual imaging protocol decisions involve clinical judgment and patient safety responsibility that cannot be delegated to unmonitored systems. Liability and regulatory coverage create strong legal barriers to full automation.
Adoption barriersclaude-sonnet-54/5Radiologic technologists must be licensed/certified, and equipment operation involving radiation exposure and patient safety carries regulatory and liability requirements that mandate human control.
Cost vs. human wageclaude-haiku-4-5-202510012/5Automation of data entry and basic documentation tasks costs less than a technician's wage, but the core task of configuring and adjusting specialized imaging hardware requires human expertise that remains cheaper to employ directly than to build custom automation infrastructure.
Cost vs. human wageclaude-sonnet-52/5Imaging equipment interfaces and physical adjustments still require a trained human operator on-site, so AI cannot substitute the full task at lower cost; software assistance saves marginal time only.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some components like image archiving and routine documentation have been partially automated, but no deployed product reliably handles the full workflow of adjusting imaging equipment parameters and selecting appropriate scan sequences without human oversight. Medical imaging software lacks autonomous capability to optimize transmitter/receiver settings.
Technical feasibility todayclaude-sonnet-52/5While voice-to-text and structured data entry tools exist, no deployed product autonomously sets scan sequences and adjusts imaging hardware without a technologist physically present and operating controls.

Complete quality control activities, monitor equipment operation, and report malfunctioning equipment to supervisor.

25

CI 25–25 · exposure 25 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare is moderately digitized but adoption of autonomous equipment monitoring remains limited; most radiology departments use human technicians for daily QC checks supplemented by vendor maintenance. Regulatory conservatism and risk aversion in medical settings slow deployment of fully autonomous monitoring.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging departments adopt AI mainly for image analysis, not equipment QC workflows; this specific administrative/maintenance task sees slow, limited AI integration.
Augmentation potentialclaude-haiku-4-5-202510013/5AI systems that flag anomalies in equipment logs, alert technicians to potential drift in calibration, or summarize maintenance alerts offer useful assistance in completing daily QC more efficiently. However, augmentation is limited to alerting and data synthesis; the technician retains decision authority.
Augmentation potentialclaude-sonnet-53/5Built-in equipment diagnostics and monitoring software can alert technologists to malfunctions and streamline reporting, improving efficiency without replacing the human role.
Task automatabilityclaude-haiku-4-5-202510012/5Monitoring equipment operation and detecting malfunctions involves pattern recognition in logs and sensor data, which AI can assist with, but quality control also requires tactile inspection, real-time judgment of image quality, and contextual interpretation of equipment behavior that demands human expertise. Complete end-to-end automation with 50% time saving at equal quality is not reliably achievable today.
Task automatabilityclaude-sonnet-52/5Quality control involves physical checks of imaging equipment, calibration tests, and hands-on inspection that current AI cannot perform autonomously; software can flag some anomalies but not execute the full physical QC workflow.
Adoption barriersclaude-haiku-4-5-202510014/5Quality control in medical imaging is subject to regulatory requirements (FDA, state licensing, accreditation standards) and liability frameworks that mandate human accountability for equipment safety and image quality assurance. A licensed radiologic technician must verify and sign off on quality control outcomes.
Adoption barriersclaude-sonnet-54/5Radiologic equipment safety and QC are subject to regulatory and accreditation requirements (e.g., ACR, state health departments) mandating human oversight and documentation, creating strong compliance barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integration of AI monitoring systems into radiology departments requires upfront infrastructure investment, ongoing maintenance, and human oversight to validate alerts. The all-in cost (licensing, integration, human review) approaches or exceeds the cost of a technician performing routine monitoring and reporting.
Cost vs. human wageclaude-sonnet-52/5Automated diagnostic monitoring exists in imaging systems but still requires human technologists on-site to interpret, act on, and physically verify equipment status, so cost savings are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI-driven monitoring systems exist for equipment diagnostics and anomaly detection in medical imaging devices, but these operate within narrow scope (specific equipment types) and typically flag concerns rather than autonomously resolve them. Deployed products require substantial human oversight and cannot reliably replace the full quality control workflow without human sign-off.
Technical feasibility todayclaude-sonnet-52/5Some equipment includes built-in diagnostic software that logs errors, but no deployed product independently performs comprehensive QC and reporting without a human technologist physically involved.

Coordinate work with clerical personnel or other technologists and technicians.

21

CI 11–30 · exposure 13 · augmentation 50 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of AI for operational coordination remains limited and slow; most hospitals still rely on manual scheduling software and human coordinators. The high stakes of patient safety and the complexity of clinical workflows have constrained experimental deployment even where technically possible.
Sector adoption velocityclaude-sonnet-52/5Healthcare administrative workflows are adopting AI scheduling and communication tools slowly compared to finance or tech sectors, with pilots more common than full deployment for staff coordination specifically.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-powered scheduling assistants and real-time messaging systems can help human coordinators track workload and flag conflicts, improving throughput and reducing coordination overhead. These tools provide useful but incremental productivity gains, not transformative augmentation of the coordinator role itself.
Augmentation potentialclaude-sonnet-53/5AI-powered scheduling, messaging, and workflow management tools can meaningfully assist technologists in coordinating tasks, though human judgment and communication remain central.
Task automatabilityclaude-haiku-4-5-202510012/5Coordination tasks involving scheduling, status updates, and information routing can be partially automated, but the interpersonal judgment and exception-handling required to manage technologist teams in real-time clinical settings resist full automation. Current AI lacks the situational awareness and negotiation capability to handle conflicts or urgent re-prioritization that human coordinators routinely manage.
Task automatabilityclaude-sonnet-51/5This is interpersonal coordination and scheduling among staff in a clinical setting, requiring real-time negotiation and physical presence that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Coordination of clinical staff sits at the intersection of patient safety regulations (HIPAA, accreditation standards) and real-time operational dependencies; healthcare institutions mandate human accountability for scheduling and resource allocation. Liability concerns and the embedded role of technologists in direct patient care create legal and organizational friction against full substitution.
Adoption barriersclaude-sonnet-53/5No licensing barrier specifically blocks AI from assisting, but organizational structure, workflow ownership, and accountability for coordinating clinical staff create practical friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Deployed scheduling and communication tools require significant integration, customization, and oversight within hospital IT systems, and they typically automate only the routine data-routing aspects. The cost of implementing and maintaining such systems often approaches or exceeds the modest clerical savings they provide, especially in smaller departments.
Cost vs. human wageclaude-sonnet-52/5Some scheduling/communication tools reduce coordination overhead cheaply, but the actual interpersonal coordination task still requires human labor, so AI alone isn't a substitute at comparable cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5While task management and scheduling software exist, they are typically passive tools that technologists use themselves rather than autonomous coordinators. No AI system reliably performs the judgment-driven parts of this task—asserting priorities, resolving resource conflicts, or adapting to emergency surgical schedules—in production healthcare settings today.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously coordinates human staff work across departments in a radiology setting; this remains a human management function.

Operate or oversee operation of radiologic or magnetic imaging equipment to produce images of the body for diagnostic purposes.

21

CI 11–30 · exposure 25 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare settings have been slow to adopt autonomous imaging operation; most AI deployment is limited to diagnostic assistance (reading support) rather than equipment operation. Regulatory conservatism, patient safety requirements, and institutional inertia constrain rapid adoption of automation in this task.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging departments are adopting AI for image analysis and workflow support, but the physical operation task itself sees minimal automation deployment.
Augmentation potentialclaude-haiku-4-5-202510014/5AI tools that suggest optimal protocols, flag quality issues in real-time, and assist with image reconstruction can meaningfully enhance technician productivity and consistency. These assistive systems are already in use and demonstrably help technicians work faster and more accurately.
Augmentation potentialclaude-sonnet-53/5AI can assist with protocol selection, image quality optimization, and workflow scheduling, improving technologist efficiency without replacing the hands-on operation.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in image analysis and quality assessment, the task requires hands-on patient positioning, equipment operation, protocol selection, and real-time safety decisions that cannot be fully automated. Current systems cannot independently manage the full workflow of patient interaction, equipment calibration, and imaging execution.
Task automatabilityclaude-sonnet-52/5Positioning patients, operating imaging hardware, adjusting for anatomy and safety requires physical manipulation and judgment that current AI cannot perform end-to-end; AI mainly assists with image analysis, not equipment operation.'
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory bodies (FDA, ACR) require licensed radiologic technologists to operate or directly oversee imaging equipment and ensure patient safety and radiation protection protocols. Legal and liability frameworks are tightly tied to human credentialing and sign-off, creating strong adoption barriers.
Adoption barriersclaude-sonnet-55/5Operating radiologic equipment on patients requires licensure, adherence to radiation safety regulations, and direct human presence, making this a hard-barrier task.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI inference for analysis is cheap, but integration into imaging workflows, validation systems, and oversight infrastructure adds significant costs. The loaded wage of a radiologic technician is modest relative to equipment and facility costs, making per-task replacement economics marginal.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing the physical equipment operation, so cost comparison favors the human technologist entirely.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI products exist for image analysis and quality control (e.g., detection aids, protocol suggestions), but no deployed system reliably performs the full operation task including patient handling and equipment oversight. Narrow-scope automation in specific imaging subtasks is feasible, but end-to-end operation remains technician-dependent.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously operates MRI/X-ray/CT equipment on patients; this remains a hands-on clinical task performed by licensed technologists.

Assign duties to radiologic staff to maintain patient flows and achieve production goals.

21

CI 11–30 · exposure 13 · augmentation 50 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare IT adoption of autonomous scheduling is slow; most radiology departments still use manual schedules or basic workforce-management software with human final approval, reflecting organizational conservatism and regulatory friction in clinical settings.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially clinical operations and imaging departments, has historically been a slower adopter of AI-driven management tools compared to sectors like finance or tech, though scheduling software adoption is increasing incrementally.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could usefully assist supervisors by suggesting staff allocations based on patient volume and technologist skills, flagging bottlenecks, and tracking real-time status, but final assignment decisions remain with human managers who integrate factors AI cannot reliably weight.
Augmentation potentialclaude-sonnet-53/5AI-based scheduling and workflow optimization tools can help supervisors visualize patient flow, predict bottlenecks, and suggest staff assignments, providing meaningful assistance while the human retains final decision authority.
Task automatabilityclaude-haiku-4-5-202510012/5Assigning staff duties requires real-time assessment of patient census, staff availability, skill levels, and dynamic workflow changes—tasks that exceed current AI systems' ability to reliably integrate multiple organizational constraints and adapt to exceptions without human oversight. While AI could assist in scheduling, the nuanced judgment needed for patient safety and production optimization remains primarily human.
Task automatabilityclaude-sonnet-51/5This is a managerial/scheduling task requiring situational judgment about staff skills, patient acuity, and workflow bottlenecks in a physical clinical setting; current AI cannot perform this end-to-end reliably.
Adoption barriersclaude-haiku-4-5-202510014/5Radiologic departments are regulated environments under HIPAA and state medical licensure rules; liability for assignment errors that affect patient safety and imaging quality rests with the supervising radiologic technologist or radiologist, creating legal and professional accountability barriers to full automation.
Adoption barriersclaude-sonnet-53/5No explicit licensing requirement for the scheduling function itself, but organizational hierarchy, accountability for clinical outcomes, and need for human judgment in dynamic clinical environments create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5An AI scheduling system would require custom integration, ongoing tuning, and human oversight (radiologic supervisors must still validate assignments), making all-in costs comparable to or exceeding the part-time coordinator role this task typically occupies.
Cost vs. human wageclaude-sonnet-52/5Some scheduling software can reduce administrative overhead, but the human judgment component (assessing staff competency, patient needs, real-time adjustments) still requires a paid supervisor, so cost savings are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5No production systems autonomously manage radiologic department staff assignments at scale; the task requires integration with EHR systems, real-time situational awareness, and accountability for patient safety that existing AI tools do not reliably handle. Scheduling software exists but does not replace human supervisor decision-making for dynamic duty assignment.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously assigns radiologic staff duties and manages patient flow in production; staffing/scheduling tools exist but require human decision-making and oversight for this specific supervisory task.

Make exposures necessary for the requested procedures, rejecting and repeating work that does not meet established standards.

19

CI 16–21 · exposure 25 · augmentation 50 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare remains a laggard sector for autonomous AI in clinical tasks; hospitals have piloted AI image-analysis aids but have not displaced technician exposure decisions. Adoption is limited to retrospective QA rather than real-time automation.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging technologist roles remain a physically-anchored, highly regulated field with slow AI adoption for the hands-on portions of the task, though AI is being piloted for image quality analytics.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by flagging common quality defects post-exposure (positioning errors, motion artifact, underexposure) and recommending parameters for repeats, moderately raising technician efficiency in quality review and decision-making.
Augmentation potentialclaude-sonnet-53/5AI-based image quality analysis tools can help technologists quickly identify exposures that don't meet standards, improving efficiency in the rejection/repeat decision process.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can help analyze image quality post-hoc, the physical act of positioning equipment, patient handling, and real-time judgment about exposure adequacy require human presence and manual adjustment. AI cannot independently perform the full end-to-end task of making exposures and deciding on-the-fly whether to repeat them.
Task automatabilityclaude-sonnet-52/5The physical positioning of patients, operating imaging equipment, and hands-on adjustment for exposures requires physical presence and manipulation that current AI cannot perform end-to-end; AI can assist in image quality assessment but not the physical act.'
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory bodies (FDA, state licensing boards) require certified radiologic technologists to perform exposures and make quality judgments; liability and safety concerns heavily restrict unsupervised automation. Licensure and direct human accountability create substantial legal barriers.
Adoption barriersclaude-sonnet-55/5Radiologic technologists must be licensed/certified, and safety regulations (radiation exposure, patient contact, equipment operation) legally require a credentialed human to perform this task.
Cost vs. human wageclaude-haiku-4-5-202510011/5Current AI support tools require significant technician oversight and do not reduce the technician's workload enough to achieve cost parity; the technician remains the primary cost driver while AI adds integration and maintenance overhead.
Cost vs. human wageclaude-sonnet-51/5AI cannot replace the physical technologist function, so cost comparison favors the human since AI alone cannot deliver the output; any AI use is supplementary and adds cost rather than replacing the wage.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI products can assess completed radiographs for quality issues, but no deployed system reliably performs independent exposure decisions, positioning, and repetition workflows in production. Research prototypes exist, but real-world radiology still requires technician judgment and intervention.
Technical feasibility todayclaude-sonnet-52/5Some AI tools exist for automated image quality checks and exposure parameter suggestions, but no deployed product autonomously performs the full exposure-taking and rejection/repeat workflow in production.

Provide students or other technicians and technologists with suggestions of additional views, alternate positioning, or improved techniques to ensure the images produced are of the highest quality.

16

CI 7–25 · exposure 13 · augmentation 50 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Radiology is highly regulated and conservative in clinical adoption; while AI-assisted image reading is advancing, the specific task of teaching technicians remains a low-priority use case; pilot programs exist but production adoption for autonomous teaching feedback is minimal.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging is a highly regulated, physically grounded sector with slower AI adoption for hands-on training and supervisory tasks compared to purely digital work.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can offer useful assistance by flagging potential quality issues, suggesting alternative angle options, or highlighting anatomy visibility—helping an experienced radiologist or senior technician explain feedback more efficiently. However, the pedagogical and judgment-intensive nature limits transformative impact.
Augmentation potentialclaude-sonnet-53/5AI image analysis tools can flag quality issues (e.g., positioning artifacts, exposure problems) that a technologist could then use to inform their feedback, offering moderate assistance.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires evaluating image quality, understanding radiologic positioning principles, and providing real-time corrective feedback tailored to specific cases—elements that demand domain expertise and contextual judgment. Current AI systems cannot reliably assess image adequacy against clinical standards or generate actionable, personalized teaching suggestions.
Task automatabilityclaude-sonnet-52/5This is a hands-on teaching and mentoring task requiring physical demonstration, judgment about positioning, and real-time feedback in a clinical environment, which current AI cannot perform end-to-end.atur
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: supervising radiologists bear professional and legal responsibility for training quality and patient safety; credentialing and institutional oversight rules typically require human experts to review and sign off on instructional feedback; liability asymmetry is high if AI-generated suggestions lead to diagnostic error.
Adoption barriersclaude-sonnet-54/5Training and quality assurance in radiologic imaging typically requires supervision by licensed, experienced technologists, creating strong organizational and credentialing barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI inference for image analysis is inexpensive, but integrating it into a teaching workflow and ensuring oversight by qualified radiologists adds significant cost; the task also demands contextual correction that a junior technician would naturally provide at low marginal cost.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this supervisory/coaching function, so no favorable cost comparison exists; a qualified human must fill this role.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can perform image quality classification in narrow, constrained benchmarks, no deployed product reliably provides teaching-grade suggestions for technique improvement in production radiology settings. Existing systems lack the pedagogical and domain-specific reasoning required for credible instruction to technicians.
Technical feasibility todayclaude-sonnet-51/5No deployed product currently supervises or trains radiologic technologists on positioning technique in clinical practice; this remains a human supervisory role.

Set up examination rooms, ensuring that all necessary equipment is ready.

14

CI 5–24 · exposure 13 · augmentation 25 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare adoption of automation for physical room setup remains minimal and limited to research pilots. Medical facilities are slow to deploy non-clinically approved automation, and technician unions and safety cultures create organizational friction against equipment-handling automation.
Sector adoption velocityclaude-sonnet-51/5Physical room setup in healthcare imaging is a low-digitization, hands-on task with no meaningful AI/robotic adoption trend in this specific sub-task.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could provide marginal assistance through inventory-tracking apps or checklists displayed on tablets, and sensor-based reminders for missing items, but these are narrow augmentations that do not substantially transform the technician's productivity on the core physical setup task.
Augmentation potentialclaude-sonnet-52/5AI could support checklist reminders or equipment status tracking via software, but this offers only marginal assistance to the core physical setup work.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist in inventory checks and equipment readiness verification through computer vision or sensor systems, the task requires physical setup, tactile configuration, and safety verification that cannot be automated end-to-end today. Current AI lacks reliable mobile manipulation and contextual judgment about proper equipment positioning.
Task automatabilityclaude-sonnet-51/5This requires physical manipulation of imaging equipment, positioning aids, and supplies in a physical room, which current AI systems cannot perform without robotic embodiment far beyond deployed capability.'
Adoption barriersclaude-haiku-4-5-202510014/5Healthcare facilities face significant regulatory requirements, liability concerns around improper equipment setup affecting patient safety, and implicit human oversight mandates. Additionally, the task touches infection control and safety protocols where human responsibility and sign-off are often legally or institutionally required.
Adoption barriersclaude-sonnet-53/5While not formally licensed as a standalone task, it is bundled into technologist duties requiring physical presence and safety compliance (e.g., radiation shielding checks), creating practical human-in-the-room requirements.
Cost vs. human wageclaude-haiku-4-5-202510011/5The capital cost of robotic systems capable of safe medical equipment handling, integration, and maintenance would far exceed the loaded wage of a radiologic technician performing manual setup. Oversight and error-correction costs would further increase total cost of ownership.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI-based alternative for this physical setup task, so AI cost is effectively infinite relative to human labor for this specific action.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed production systems currently perform full examination-room setup autonomously. Robotic arms and autonomous systems exist in research settings but lack the flexibility, safety assurance, and integration with medical facility workflows needed for reliable real-world deployment in healthcare environments.
Technical feasibility todayclaude-sonnet-51/5No deployed product physically sets up radiology exam rooms; this remains entirely a manual, hands-on task performed by technologists.

Assist with on-the-job training of new employees or students or provide input to supervisors regarding training performance.

14

CI 7–21 · exposure 5 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare organizations, especially in hospital radiology departments, adopt training AI slowly; most still rely on senior technologists and formal didactic programs, with digital tools used only for supplementary materials rather than primary training or performance assessment.
Sector adoption velocityclaude-sonnet-52/5Healthcare training and supervisory functions see slow AI adoption due to the interpersonal, hands-on nature of clinical mentorship, despite AI's growing role in radiology image analysis.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully augment trainers by generating reference materials, flagging common procedural errors from recorded sessions, or automating quiz/knowledge checks, thereby reducing some administrative burden while trainers remain essential for hands-on guidance and safety oversight.
Augmentation potentialclaude-sonnet-53/5AI tools like e-learning modules, simulation software, or performance tracking dashboards can support training documentation and provide supplementary materials, aiding but not replacing the human mentor.
Task automatabilityclaude-haiku-4-5-202510011/5Training new employees requires real-time interaction, adaptive feedback, and judgment about individual learning needs—tasks that current AI systems cannot perform end-to-end in clinical or hands-on settings without substantial human oversight.
Task automatabilityclaude-sonnet-51/5On-the-job training and performance evaluation of trainees requires hands-on demonstration, live feedback during patient interactions, and interpersonal judgment that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Clinical training of radiologic technologists is tightly regulated (accreditation, competency assessment, radiation safety certification); supervisory oversight of training performance and sign-off on competency typically require credentialed human judgment, creating a strong legal and regulatory barrier to full automation.
Adoption barriersclaude-sonnet-53/5While not formally licensed as a distinct task, training and evaluating staff involves organizational trust, accountability, and interpersonal judgment that create moderate structural resistance to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI training tools exist but typically require setup, customization, and extensive human validation; the all-in cost of implementing and monitoring AI-assisted training often exceeds the cost of a trainer's time for small cohorts in specialized technical fields.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for this task, so cost comparison favors the human trainer entirely; any AI attempt would add cost without replacing the function.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate training materials or provide asynchronous feedback on written assignments, no deployed product reliably handles the interactive mentoring, safety supervision, and performance assessment required for on-the-job radiologic technologist training in real hospital environments.
Technical feasibility todayclaude-sonnet-51/5No deployed AI products conduct hands-on clinical training or evaluate trainee performance in radiologic technology settings; this remains a purely human supervisory function.

Monitor patients' conditions and reactions, reporting abnormal signs to physician.

14

CI 3–25 · exposure 13 · augmentation 50 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While some radiology departments pilot AI-assisted imaging analysis, real-time patient monitoring and condition assessment remain primarily human functions with slow organizational adoption of autonomous alternatives. Adoption is limited by regulatory requirements, liability concerns, and lack of proven autonomous systems.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging settings adopt AI mainly for image analysis and workflow, not for real-time patient monitoring and clinical reporting, which remains a slow-adoption area due to safety-critical nature.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist technologists by flagging abnormal imaging findings and alerting them to potential concerns, raising their efficiency in triage and documentation. However, the assistive gain is moderate because the core task of clinical observation and communication still heavily relies on human judgment and direct patient interaction.
Augmentation potentialclaude-sonnet-53/5AI-enabled monitors (e.g., vital sign sensors with alerting) can flag abnormal readings to assist the technologist, improving vigilance without replacing human judgment.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI cannot reliably monitor ongoing patient conditions and reactions in real-time across the full spectrum of abnormalities, nor can it autonomously decide which signs warrant physician notification without high false-positive rates. While AI can flag certain imaging findings, the broader patient monitoring and clinical judgment required remain substantially human responsibilities.
Task automatabilityclaude-sonnet-51/5This requires real-time physical presence, hands-on patient observation, and clinical judgment about a live patient's condition during a procedure, which current AI cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Radiologic technologists are licensed professionals, and patient monitoring is a safety-critical task with regulatory oversight (state licensure, facility compliance, liability for missed findings). Clinical standards and malpractice risk create substantial legal and organizational barriers to full automation without human accountability.
Adoption barriersclaude-sonnet-55/5Patient monitoring during medical procedures requires a licensed technologist physically present, with legal and safety liability strongly tied to human accountability.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-based monitoring systems require significant infrastructure, oversight, and integration costs, and they still require human radiologic technologists to perform the core monitoring task. The cost per task-equivalent remains comparable to or exceeds human labor when fully accounting for setup, maintenance, and necessary human oversight.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this full task, so cost comparison favors the human technologist who is already present and required.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some AI systems exist for specific imaging analysis and alert flags in radiology workflows, but they lack the contextual awareness and real-time monitoring capability to handle the full task of patient observation and clinical decision-making reliably in production settings. Existing solutions are narrow and often require human validation.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously monitors patients during radiologic procedures and reports abnormal signs to physicians; monitoring systems exist but require human interpretation and action in this context.

Provide assistance to physicians or other technologists in the performance of more complex procedures.

14

CI 3–25 · exposure 13 · augmentation 38 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare radiology is adopting AI for image analysis and diagnosis, but adoption of procedural-assistance automation remains minimal; regulatory caution, liability concerns, and the need for physical presence mean the sector is moving slowly on this specific task.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging support roles show slow AI adoption for physical tasks, though imaging interpretation itself sees faster AI uptake; this specific hands-on assistance task lags.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by flagging abnormalities, providing protocol reminders, or organizing imaging data before complex procedures, moderately improving technologist efficiency, but the core assistance role remains human-centered and cannot be fully transformed by current AI.
Augmentation potentialclaude-sonnet-52/5AI can assist with workflow scheduling, equipment calibration alerts, or documentation during procedures, but offers minimal direct enhancement to the physical assistance task itself.
Task automatabilityclaude-haiku-4-5-202510012/5This task involves real-time procedural assistance, decision-making about when to adjust positioning or technique, and contextual judgment about patient safety. Current AI systems can provide some preparation or documentation support, but cannot reliably take the active-assistance role without human oversight, and automation would not achieve 50% time savings at equal quality.
Task automatabilityclaude-sonnet-51/5This task involves hands-on physical assistance during medical imaging procedures (positioning patients, handling equipment, sterile technique) which requires physical presence and manual dexterity that current AI cannot replicate.
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory bodies (FDA, state licensing boards) impose strict requirements on who can assist in complex radiologic procedures; patient safety liability and malpractice concerns create strong legal barriers to substituting human judgment with automated systems in this clinical setting.
Adoption barriersclaude-sonnet-55/5Requires licensed, certified radiologic technologists physically present for procedures involving patient safety, radiation exposure, and regulatory compliance in clinical settings.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of developing, validating, and overseeing an AI system to safely assist in complex procedures would exceed the loaded wage of a technologist providing that assistance; the specialized domain, liability exposure, and customization required make this economically unfavorable for automation.
Cost vs. human wageclaude-sonnet-51/5There is no AI system that substitutes for this physical assistance role, so cost comparison favors the human by default since no viable AI alternative exists.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can analyze images and provide diagnostic suggestions, no deployed product today reliably performs procedural assistance—the task requires physical presence, real-time adaptation, and safety-critical decisions that are not yet automated in production radiology workflows.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs physical procedural assistance to physicians in radiology suites; this remains firmly in the human physical labor domain.

Perform procedures, such as linear tomography, mammography, sonograms, joint and cyst aspirations, routine contrast studies, routine fluoroscopy, or examinations of the head, trunk, or extremities under supervision of physician.

9

CI 3–16 · exposure 13 · augmentation 63 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Despite widespread adoption of AI image-reading assistants in radiology departments, procedural automation adoption in clinical settings remains minimal. Technician displacement through full procedure automation is rare; most adoption is augmentative (assisting interpretation) rather than substitutive.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging technical labor is a physically-grounded, highly regulated sector with slow adoption of automation for hands-on procedures, though AI adoption in image interpretation is growing separately.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist technicians by providing real-time image quality feedback, protocol guidance, positioning recommendations, and post-procedure image analysis, substantially raising throughput and consistency. Technicians remain in the loop for all hands-on procedural steps while benefiting from decision support.
Augmentation potentialclaude-sonnet-53/5AI can assist with image quality optimization, positioning guidance, and post-acquisition image analysis, improving efficiency, but the core procedural execution remains manual.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with image analysis and post-processing, the task involves hands-on procedural execution (positioning patients, operating imaging equipment, performing aspirations) that requires physical manipulation and real-time clinical judgment under physician supervision. Current AI cannot independently perform the procedural components that constitute the bulk of this task.
Task automatabilityclaude-sonnet-51/5This requires physical patient handling, precise equipment positioning, sterile procedures, and hands-on manipulation (e.g., joint/cyst aspirations) that current AI cannot perform without a robotic embodiment, which is not commercially deployed for this purpose.
Adoption barriersclaude-haiku-4-5-202510015/5Strong regulatory and legal barriers exist: procedures like joint and cyst aspirations are invasive, patient safety and informed consent requirements mandate human presence, medical licensing requirements apply to procedure oversight, and liability asymmetry means errors in positioning or technique carry clinical risk that cannot be fully transferred to AI.
Adoption barriersclaude-sonnet-55/5Licensed radiologic technologists and physician supervision are legally mandated for these invasive and diagnostic procedures, and direct physical patient contact is required by regulation and safety standards.
Cost vs. human wageclaude-haiku-4-5-202510011/5The procedural and hands-on nature of imaging work means that even advanced AI for image interpretation cannot substitute for the technician's wage without substantial infrastructure change. The integrated cost of any automation would not yet undercut skilled technician labor for the complete task.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing the physical procedure itself, so cost comparison favors the human technologist entirely; AI cannot replace the physical labor component.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI systems exist for image interpretation and analysis, but deployed products do not reliably perform the full procedural workflow—patient positioning, equipment operation, real-time guidance, and aspirations—that defines this occupational task. Clinical deployment focuses on post-hoc analysis rather than procedure execution.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously performs imaging procedures or aspirations on patients; this remains purely a human physical task with AI only assisting in image analysis after acquisition.

Position imaging equipment and adjust controls to set exposure time and distance, according to specification of examination.

7

CI 0–14 · exposure 8 · augmentation 38 · importance 5.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a laggard sector in adopting physical automation for equipment operation, with high regulatory caution, strong preference for licensed human control, and limited capital deployment in robotic positioning systems for diagnostic imaging.
Sector adoption velocityclaude-sonnet-51/5Healthcare imaging technician roles involve physical, hands-on work in a highly regulated, slow-to-digitize physical environment, showing minimal automation deployment for this specific task.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by recommending exposure parameters based on patient demographics and examination type, and by alerting technicians to positioning errors, but the technician remains essential for safe hands-on equipment adjustment and patient interaction.
Augmentation potentialclaude-sonnet-52/5AI can suggest optimal exposure parameters or protocols based on exam type, offering some decision support, but does not meaningfully assist the physical positioning process itself.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can theoretically assist with calculating exposure parameters, the physical positioning of equipment and real-time adjustment based on patient anatomy requires robotic hardware integration and reliable real-world perception—capabilities not yet deployed at scale in clinical settings. Current systems lack the embodied dexterity and safety-critical reliability needed for end-to-end automation.
Task automatabilityclaude-sonnet-51/5This requires physically positioning equipment and patients, and adjusting hardware controls in a real clinical room—current AI systems cannot perform physical manipulation or patient handling.'
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory bodies (FDA, state radiologic licensing boards) require qualified, licensed radiologic technologists to perform or directly supervise imaging operations due to radiation safety and patient safety mandates. Liability for equipment misalignment or incorrect exposure parameters creates strong legal barriers to full automation.
Adoption barriersclaude-sonnet-55/5Radiologic technologists must be licensed/certified, and positioning patients near radiation sources involves direct physical patient contact and safety/regulatory requirements that mandate human oversight.
Cost vs. human wageclaude-haiku-4-5-202510011/5Robotic systems capable of equipment positioning and control adjustment are expensive to purchase, maintain, and integrate, far exceeding the hourly cost of a technician for this specific task component.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical task, so cost comparison favors the human by default since no functioning AI alternative exists.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed products today reliably perform the full task of physically positioning imaging equipment and adjusting controls autonomously in a clinical environment. Research prototypes exist, but production systems performing this consistently in hospitals do not.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously positions imaging equipment or physically sets up patients for exposure; this remains entirely a hands-on human task.

Position patient on examining table and set up and adjust equipment to obtain optimum view of specific body area as requested by physician.

7

CI 0–14 · exposure 13 · augmentation 38 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare settings, particularly imaging departments, have adopted AI narrowly for image interpretation and protocol support, not for physical task automation. Adoption of autonomous patient positioning remains at pilot/research stage with minimal production deployment in real clinical settings.
Sector adoption velocityclaude-sonnet-51/5Healthcare physical-task automation, especially direct patient contact and equipment handling, is a laggard area with minimal real-world deployment of robotic or AI-driven positioning systems.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist through automated protocol recommendations, real-time feedback on positioning adequacy relative to diagnostic standards, and image-quality guidance—raising technologist efficiency in executing the positioning task while the technologist retains control and ensures patient safety.
Augmentation potentialclaude-sonnet-52/5AI can assist with equipment settings suggestions or protocol optimization software, but it offers little assistance for the core physical act of positioning patients and adjusting hardware.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist with equipment parameter optimization based on imaging protocols, the physical task of patient positioning requires human judgment, safety awareness, and real-time adaptation to patient condition—elements current AI systems cannot perform autonomously. Meaningful automation is limited to software-driven protocol recommendations, not the full end-to-end task.
Task automatabilityclaude-sonnet-51/5This is a physical, hands-on task requiring manual patient positioning and physical equipment adjustment; no current AI system can perform physical manipulation of patients or imaging hardware.
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and safety barriers exist: patient safety protocols, radiologic protection standards, and direct patient contact requirements mandate human oversight. Liability for positioning errors (injury, poor image quality) creates asymmetric error costs that discourage full automation without licensed human supervision.
Adoption barriersclaude-sonnet-55/5Patient handling and positioning for diagnostic imaging is subject to licensure requirements, safety regulations, and direct human-contact/liability considerations that legally require a trained technologist.
Cost vs. human wageclaude-haiku-4-5-202510011/5The capital cost of robotic systems capable of safe patient positioning, combined with integration and oversight, would substantially exceed the loaded wage of a trained radiologic technologist. Current automation solutions do not achieve cost parity for this task.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical task, so any hypothetical robotic solution would require expensive specialized hardware far exceeding current human labor costs.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs autonomous patient positioning and equipment setup. Computer-aided detection systems exist for image analysis but not for the physical manipulation and real-time adjustment required by this task. Clinical deployment would require solving safety and liability challenges that remain unsolved.
Technical feasibility todayclaude-sonnet-51/5No deployed product physically positions patients or adjusts imaging equipment autonomously today; this remains entirely a research/robotics-frontier concept, not a production capability.

Prepare contrast material, radiopharmaceuticals, or anesthetic or antispasmodic drugs under the direction of a radiologist.

6

CI 0–11 · exposure 8 · augmentation 38 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Radiology departments are digitizing imaging and reporting, but pharmaceutical preparation remains a manual, regulated process with slow AI adoption due to safety-critical requirements and resistance to automation in controlled pharmacy environments.
Sector adoption velocityclaude-sonnet-51/5Physical pharmaceutical preparation in clinical radiology settings shows negligible AI adoption; this remains a hands-on, low-digitization task.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by automating dosage calculation verification, flagging drug interactions, managing inventory and expiration tracking, and documenting preparation steps—improving workflow efficiency while the technician retains responsibility for safe handling.
Augmentation potentialclaude-sonnet-52/5AI could assist with dosage calculations, protocol lookups, or documentation, but offers minimal help with the physical mixing and preparation steps themselves.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist with inventory management and documentation, the physical preparation of contrast agents and radiopharmaceuticals, measurement accuracy, and aseptic handling require human manipulation and judgment. Current systems cannot reliably handle the manual procedural steps or adapt to real-time complications without supervision.
Task automatabilityclaude-sonnet-51/5This is a manual, physical task requiring precise handling, mixing, and administration of pharmaceuticals and contrast agents; no current AI system can physically prepare or handle these substances.
Adoption barriersclaude-haiku-4-5-202510015/5Healthcare regulations, pharmacy licensing requirements, and clinical protocol mandates mean a licensed radiologic technologist or pharmacist must legally prepare and verify these materials; liability for pharmaceutical errors is high and non-delegable to automated systems.
Adoption barriersclaude-sonnet-55/5Preparation of contrast media, radiopharmaceuticals, and controlled anesthetic/antispasmodic drugs is tightly regulated and requires licensed, trained personnel working under physician direction, with strict liability and safety requirements.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of AI infrastructure, oversight, and quality assurance to manage pharmaceutical preparation would exceed the labor cost of a trained technician, especially given liability and compliance requirements in healthcare settings.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical preparation task, so cost comparison is moot—human labor is the only option at present.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI products perform end-to-end preparation of pharmaceutical materials; this task involves physical chemistry, aseptic technique, and regulatory compliance that require human execution. AI tools for documentation exist but cannot substitute for the actual preparation work.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical drug/contrast preparation; this remains entirely a human manual and clinical task requiring hands-on execution.

Explain procedures and observe patients to ensure safety and comfort during scan.

4

CI 0–7 · exposure 0 · augmentation 25 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Patient-facing clinical safety tasks in radiology are subject to regulatory constraints and institutional friction that strongly protect human technician roles. Adoption of automation in this domain is minimal and unlikely to accelerate.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging settings adopt AI for image analysis but adoption of patient-facing physical monitoring/interaction tasks remains minimal and slow.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could minimally assist with explaining standardized procedure protocols or generating comfort recommendations, but the core task of real-time patient observation and reassurance depends on human presence and cannot be augmented meaningfully by automation.
Augmentation potentialclaude-sonnet-52/5AI-driven patient education videos or chatbots could supplement pre-scan explanations, but real-time comfort/safety observation during the scan itself sees little AI assistance.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires direct human-patient interaction, real-time observation of patient condition, and dynamic safety adjustments that current AI systems cannot perform autonomously. Patient reassurance and comfort assessment are inherently relational and require human presence.
Task automatabilityclaude-sonnet-51/5This requires physical presence, hands-on positioning, real-time observation of patient distress, and physical safety interventions that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Strict regulatory and legal requirements mandate that a licensed technician physically present and directly responsible for patient safety, comfort, and procedure compliance during imaging scans. Healthcare liability and patient care standards create hard barriers to substitution.
Adoption barriersclaude-sonnet-54/5Direct patient contact, safety monitoring, and procedural explanation typically require a licensed technologist present, driven by liability, patient safety regulations, and accreditation standards.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems cannot replace the human technician's physical presence, real-time monitoring, and safety oversight. The cost of deployment would exceed the wage value of the task.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical, in-person task, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs real-time patient observation, safety monitoring, and comfort management during medical imaging. This remains a human-centered clinical task with no AI alternative in production.
Technical feasibility todayclaude-sonnet-51/5No deployed product physically monitors and reassures patients during imaging procedures or handles their physical safety in the scan room.

Transport patients to or from exam rooms.

3

CI 0–5 · exposure 0 · augmentation 13 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a laggard sector for this task; no measurable production deployment of autonomous patient transport exists in clinical settings.
Sector adoption velocityclaude-sonnet-51/5Healthcare physical-task settings show minimal AI/robotic adoption for patient transport; this remains a manual, human-performed task with negligible automation penetration.
Augmentation potentialclaude-haiku-4-5-202510012/5Limited augmentation possible; scheduling software or navigation aids could assist logistics minimally, but the core task demands human physical presence and attention.
Augmentation potentialclaude-sonnet-51/5AI offers essentially no assistance for the physical act of moving patients between rooms; this is a manual logistics/physical care task outside AI's current capabilities.
Task automatabilityclaude-haiku-4-5-202510011/5Transporting patients to and from exam rooms requires physical movement, human judgment about patient safety/mobility, and interpersonal care. Current AI systems cannot physically manipulate or move people reliably.
Task automatabilityclaude-sonnet-51/5Physically transporting patients (wheelchair/gurney transfers) requires physical robotics manipulation of people, which current AI systems cannot perform; this is a physical mobility task, not cognitive.rating reflects no viable AI substitute today.
Adoption barriersclaude-haiku-4-5-202510015/5Patient safety, liability, and duty of care create hard barriers; direct human contact and judgment are legally and medically required for safe transport of vulnerable individuals.
Adoption barriersclaude-sonnet-54/5Patient handling involves safety, liability, and physical care considerations often requiring trained personnel, and healthcare facilities have strict protocols around patient movement and safety, creating strong organizational and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Mobile robotics for patient transport are experimental and prohibitively expensive compared to the loaded wage of a technician or assistant.
Cost vs. human wageclaude-sonnet-51/5There is no functioning AI/robotic system performing this task at scale, so cost comparison favors the human worker entirely; any robotic alternative would require expensive specialized hardware exceeding labor costs.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs patient transport; this remains purely a human task in medical settings due to safety, liability, and physical contact requirements.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product moves or transfers patients between exam rooms; autonomous patient transport robots remain experimental/research-stage, not standard hospital equipment.

Provide assistance in dressing or changing seriously ill or injured patients or patients with disabilities.

3

CI 0–5 · exposure 0 · augmentation 13 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a laggard sector for physical automation; adoption of robotic assistance for patient handling is minimal and largely confined to research settings or specialized facilities, not mainstream clinical practice.
Sector adoption velocityclaude-sonnet-51/5Physical patient care tasks in healthcare settings show very low AI/robotic adoption due to safety, dexterity, and regulatory constraints.
Augmentation potentialclaude-haiku-4-5-202510012/5While minor AI-driven aids (alerts, reminders about patient mobility status) could assist human technicians, AI does not meaningfully augment the core physical task of dressing or changing patients, which relies on direct tactile interaction and human judgment about patient comfort.
Augmentation potentialclaude-sonnet-51/5AI offers essentially no meaningful assistance for the physical act of dressing or repositioning patients; this remains a fully manual task.
Task automatabilityclaude-haiku-4-5-202510011/5This task involves direct physical contact and manipulation of patients' clothing and bodies, which requires embodied dexterity, force calibration, and sensitivity to patient comfort and safety that current AI systems cannot perform. No robotic or AI system can reliably handle the physical and interpersonal complexity of dressing seriously ill patients.
Task automatabilityclaude-sonnet-51/5This is a hands-on physical caregiving task requiring manipulation of a patient's body, clothing, and medical equipment (IVs, wound dressings); no current AI system can perform physical manipulation of this kind.directly, robotics for this remain research-stage.
Adoption barriersclaude-haiku-4-5-202510015/5This task involves direct physical contact with vulnerable patients and is legally and professionally the responsibility of licensed medical personnel. Healthcare liability, patient safety regulations, and institutional policy require a qualified human to perform this task, creating hard legal and organizational barriers to substitution.
Adoption barriersclaude-sonnet-54/5Direct physical patient care involving disabled or seriously ill individuals carries significant liability, safety, and dignity concerns, generally requiring trained human staff, though not always a licensed technologist specifically.
Cost vs. human wageclaude-haiku-4-5-202510011/5The capital cost of any robotic system capable of this task (humanoid or specialized manipulation robot) combined with integration, maintenance, and reliability oversight would far exceed the loaded wage of a radiologic technologist or nursing assistant performing this work.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute, so any AI-based approach (e.g., robotic assistance) would be far more costly than a human aide performing this task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI or robotic system reliably performs this task in production healthcare environments today. While research exists in robotic manipulation, nothing at scale performs end-to-end patient dressing assistance in real clinical settings.
Technical feasibility todayclaude-sonnet-51/5No deployed AI or robotic product performs patient dressing/changing assistance in clinical settings; this remains purely a human physical care task.

Use radiation safety measures and protection devices to comply with government regulations and to ensure safety of patients and staff.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare settings remain highly regulated with strong organizational resistance to removing human oversight of safety-critical tasks. Adoption of AI for documentation support is slow, and automation of actual safety enforcement is virtually absent in production healthcare systems.
Sector adoption velocityclaude-sonnet-52/5Healthcare imaging is a highly regulated, physically-grounded sector where AI adoption for safety-critical hands-on tasks remains slow despite AI's growing use in image analysis.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by logging safety checklists, flagging missing documentation, or reminding staff of protocol steps, but the human technologist must remain the decision-maker and executor of all actual protective measures and compliance verification.
Augmentation potentialclaude-sonnet-52/5AI can support dose-tracking software or alert systems that help technologists monitor radiation exposure, but it plays a minor role compared to human execution of safety protocols.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time physical enforcement of safety protocols, positioning of protective equipment, and situational judgment about radiation hazards in clinical settings. AI cannot physically place lead aprons, adjust shielding, or monitor live staff behavior in exam rooms.
Task automatabilityclaude-sonnet-51/5This task requires physically positioning shielding, operating equipment controls, and making real-time judgment calls with patients in a clinical setting, none of which current AI can perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Government regulations (NRC, state health departments) explicitly require a licensed radiologic technologist to implement and verify radiation safety measures. Legal liability for patient/staff harm creates a hard requirement for human professional accountability and sign-off.
Adoption barriersclaude-sonnet-55/5Radiation safety compliance is governed by strict government regulations (e.g., ALARA, licensing bodies) requiring certified human technologists to be physically present and accountable, creating hard legal and physical barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems capable of monitoring safety compliance would require extensive hardware integration (cameras, sensors) and human oversight to verify correct implementation, making the all-in cost of AI-assisted enforcement comparable to or exceeding direct human supervision.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical safety task, so no cost comparison favors AI; a human technologist is required regardless of cost.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system can autonomously enforce or implement physical radiation safety measures. While AI can assist in documentation and compliance tracking, the core task—actual deployment and supervision of protection devices—remains entirely human-dependent.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product physically applies lead shielding, sets exposure parameters, or manages hands-on patient safety protocols; this remains entirely a human physical and regulatory task.

Operate mobile x-ray equipment in operating room, emergency room, or at patient's bedside.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a laggard sector for automation of clinical equipment operation, with strong human-contact requirements, regulatory oversight, and low organizational push toward removing technologists from patient-facing roles.
Sector adoption velocityclaude-sonnet-51/5Physical, hands-on healthcare tasks involving patient contact and mobile equipment show minimal AI/robotic adoption; this is a low-digitization, high-physical-presence task.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with image analysis and positioning recommendations after capture, but current systems offer minimal real-time assistance during the physical equipment operation itself, which remains technologist-centric.
Augmentation potentialclaude-sonnet-52/5AI can assist with image quality optimization or post-capture image analysis, but offers little assistance with the physical operation, positioning, and equipment transport aspects of this specific task.
Task automatabilityclaude-haiku-4-5-202510011/5Operating mobile x-ray equipment requires physical manipulation of heavy equipment, real-time positioning at a patient's bedside, and immediate human interaction in clinical settings. Current AI cannot perform the hands-on equipment operation, spatial navigation, or patient handling required.
Task automatabilityclaude-sonnet-51/5This requires physically transporting equipment, positioning patients (often unconscious, injured, or in sterile OR environments), and manipulating machinery hands-on—none of which current AI systems can perform without embodied robotics far beyond present deployment.
Adoption barriersclaude-haiku-4-5-202510015/5Strict regulatory requirements (FDA, state licensing boards) mandate that qualified radiologic technologists perform x-ray operations; medical liability, equipment safety protocols, and direct patient contact create substantial legal and organizational barriers to automation.
Adoption barriersclaude-sonnet-55/5Requires licensed radiologic technologists, direct physical patient contact, sterile/emergency protocols, and legal/regulatory requirements for imaging personnel, creating hard barriers to any automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The hardware and infrastructure costs to automate mobile x-ray equipment operation (robotics, real-time navigation, safety systems) far exceed the loaded wage of a radiologic technologist, with no commercially viable solution available.
Cost vs. human wageclaude-sonnet-51/5There is no AI system performing this physical task, so cost comparison favors the human by default; robotic alternatives would be far more expensive than technician wages even if they existed.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system currently operates mobile x-ray equipment autonomously. This task requires precise physical control in dynamic clinical environments where human presence is mandatory for patient safety and regulatory compliance.
Technical feasibility todayclaude-sonnet-51/5No deployed products autonomously operate mobile x-ray equipment at bedside or in ORs; this remains firmly a human physical task with no robotic substitute in clinical use.

Operate fluoroscope to aid physician to view and guide wire or catheter through blood vessels to area of interest.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of autonomous or near-autonomous fluoroscope operation is virtually nonexistent. Healthcare and medical imaging sectors remain conservative with safety-critical procedural automation, and no market-driven displacement is evident.
Sector adoption velocityclaude-sonnet-51/5Interventional radiology and cath lab settings are highly physical, safety-regulated environments with minimal AI-driven displacement of hands-on equipment operation to date.
Augmentation potentialclaude-haiku-4-5-202510012/5Current AI can assist with image analysis or recommend positioning, but the real-time control and coordination demands mean augmentation remains limited. The technician retains primary manual control and decision-making throughout the procedure.
Augmentation potentialclaude-sonnet-53/5AI-enhanced image processing, noise reduction, and real-time image guidance software can improve visualization quality and reduce radiation dose, aiding the technologist and physician during the procedure.
Task automatabilityclaude-haiku-4-5-202510011/5Operating a fluoroscope to guide catheters requires real-time, safety-critical manual control and live interaction with a physician during a procedure. Current AI systems cannot reliably handle the coordination, physical manipulation, and dynamic decision-making needed in this clinical context.
Task automatabilityclaude-sonnet-51/5This is a hands-on, real-time procedural task requiring physical positioning of imaging equipment and coordination with a physician during an invasive procedure; no current AI system can perform this end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5This task is firmly under medical licensing and liability law: a licensed radiologic technologist must operate the equipment and bear responsibility for patient safety. Regulatory frameworks (FDA, state licensure) legally require human credentialed personnel for fluoroscopy operations.
Adoption barriersclaude-sonnet-55/5This involves operating radiation-emitting equipment during an invasive procedure, requiring licensure, direct physician collaboration, and strict regulatory/safety oversight, making autonomous substitution legally and physically barred.
Cost vs. human wageclaude-haiku-4-5-202510011/5The equipment, regulatory compliance, and integration costs far exceed the savings from AI assistance in this safety-critical medical procedure. Human technicians are necessary for immediate oversight and adjustment.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical, safety-critical function, so any AI cost comparison is moot; the human technologist remains the only cost-effective option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product currently operates fluoroscopes autonomously or guides catheter placement in production. This remains a physician-led task requiring licensed human technicians; no commercial system performs this reliably without direct human control.
Technical feasibility todayclaude-sonnet-51/5No deployed product operates fluoroscopy equipment autonomously to guide catheters through vasculature; this remains far outside current product capability, existing only in early robotic-assisted research contexts for related but distinct roles.

Related occupations — Healthcare Practitioners & Technical

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.