Diagnostic Medical Sonographers
29-2032.00Median wage $96,590/yr90,160 employed (US)Rank #584 of 923 scored · top 63% by substitution
Produce ultrasonic recordings of internal organs for use by physicians. Includes vascular technologists.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution24
Exposure23
Augmentation46
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
20 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
5%
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%24
panel mean rating 1.9/5 → substitution pressure 24/100
Technical feasibility todayw 20%22
panel mean rating 1.9/5 → substitution pressure 22/100
Cost vs. human wagew 15%20
panel mean rating 1.8/5 → substitution pressure 20/100
Adoption barriersw 20%inverted — strong barriers lower the score27
panel mean rating 3.9/5 (barrier strength) → substitution pressure 27/100
Sector adoption velocityw 10%25
panel mean rating 2.0/5 → substitution pressure 25/100
Task breakdown (20 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.
Perform clerical duties, such as scheduling exams or special procedures, keeping records, or archiving computerized images.
85CI 75–95 · exposure 87 · augmentation 75 · importance 4.1/5 · click for rater detail
Perform clerical duties, such as scheduling exams or special procedures, keeping records, or archiving computerized images.
85| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Healthcare organizations, especially imaging departments, have already widely adopted automated scheduling systems, EHR integration, and PACS with AI features. Adoption is mature and ongoing in the information-rich healthcare sector. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Healthcare administrative software adoption is fast-moving, with scheduling and EHR/PACS integration now standard in most hospital and imaging center IT stacks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI scheduling assistants and automated archival systems consistently improve staff productivity by reducing manual data entry and search time, allowing sonographers to focus on clinical work while systems handle routine administrative overhead. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-driven scheduling assistants and automated archiving substantially reduce sonographers' administrative burden, letting them focus more on scanning and patient care. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Scheduling exams, record-keeping, and archiving digital images are core clerical tasks with clear business logic and standardized workflows. Current AI systems can handle appointment scheduling, maintain patient records, and manage digital asset organization end-to-end, easily meeting the 50% time-saving bar. |
| Task automatability | claude-sonnet-5 | 4/5 | Scheduling, records management, and image archiving are largely structured digital workflows well-suited to existing scheduling software, EHR automation, and PACS systems with AI-assisted indexing. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While HIPAA and data governance impose compliance requirements on how systems handle health records, these barriers apply to the data, not to the legal permissibility of AI automation. No licensing requirement mandates a human perform scheduling or archiving; oversight is organizational rather than regulatory. |
| Adoption barriers | claude-sonnet-5 | 2/5 | These are clerical, non-clinical tasks with no licensure requirement, though healthcare-specific compliance (HIPAA, data retention rules) adds some integration friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Scheduling and archival automation costs (cloud storage, integrated scheduling APIs, record management systems) are substantially cheaper than paying full-time clerical staff, often by an order of magnitude when amortized per task. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated scheduling and archiving software costs a small fraction of dedicating clinical or clerical staff time to these administrative tasks, especially at scale. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Mature production systems already perform these functions: appointment scheduling software, EHR systems with AI-assisted documentation, and PACS (picture archiving systems) with automated indexing and retrieval are deployed and reliable in healthcare organizations today. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Practice management systems, EHR scheduling modules, and automated PACS archiving are already deployed at scale in most imaging departments, though some manual coordination and exception handling remains. |
Maintain stock and supplies, preparing supplies for special examinations and ordering supplies when necessary.
61CI 48–75 · exposure 62 · augmentation 63 · importance 3.9/5 · click for rater detail
Maintain stock and supplies, preparing supplies for special examinations and ordering supplies when necessary.
61| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare is digitizing supply chains and inventory management rapidly across major systems; hospital informatics and supply-chain automation are mainstream, with widespread EHR and materials-management system adoption reported in large and mid-size facilities. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administrative/logistics functions adopt digital tools slowly relative to information-sector benchmarks, though inventory software itself is mature and common. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools can assist sonographers by alerting to low stock, recommending reorders based on demand forecasts, and streamlining supply preparation workflows; these augmentations reduce cognitive load and free time for clinical duties while sonographers remain in the loop for final approval. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-assisted inventory tracking and automated reorder alerts can meaningfully reduce manual monitoring effort, but the physical preparation of exam-specific supplies still requires the sonographer. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | This task involves routine inventory tracking, ordering, and preparation that can be largely automated: inventory monitoring via sensors/systems, automated reordering based on thresholds, and supply staging could all be handled by AI-integrated systems with minimal human oversight, easily achieving 50% time savings. |
| Task automatability | claude-sonnet-5 | 3/5 | Inventory tracking and reordering are well-suited to automation with existing inventory management software, but physically preparing supplies for exams still requires human handling. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While supply management is less regulated than clinical work, some hospitals require human sign-off on certain orders, and integration with procurement and material-management systems introduces organizational friction; no strict licensing barrier exists, but workflow inertia is moderate. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for supply management itself, though hospital procurement rules and physical presence for exam prep create minor organizational friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Inventory automation software and supply-chain integration have low per-transaction costs once deployed, and healthcare systems already leverage these widely; the cost per 'ordering event' is dramatically lower than the time a skilled sonographer would spend, likely an order of magnitude difference. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Basic inventory software is cheap and widely used, but integration with hospital procurement systems and human oversight for special exam prep keeps costs comparable to current practice rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed inventory management systems and automated supply-chain tools reliably handle stock monitoring and reordering in healthcare settings; integrating these with sonography-specific workflows is straightforward and already occurring in practice, though some manual verification remains. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Automated inventory/reorder systems are common in healthcare settings, but they're typically generic supply-chain tools rather than AI systems built specifically for this task, and physical prep is unautomated. |
Maintain records that include patient information, sonographs and interpretations, files of correspondence, publications and regulations, or quality assurance records, such as pathology, biopsy, or post-operative reports.
60CI 50–70 · exposure 62 · augmentation 88 · importance 4.6/5 · click for rater detail
Maintain records that include patient information, sonographs and interpretations, files of correspondence, publications and regulations, or quality assurance records, such as pathology, biopsy, or post-operative reports.
60| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare IT adoption is accelerating, with many large hospital networks and imaging centers already deploying AI-assisted documentation and records management as part of digital transformation initiatives. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare documentation systems are adopting AI tools like natural language processing and automated report generation, but overall sector adoption remains uneven and pilot-heavy due to regulatory and interoperability constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI tools like automated transcription, template-based note generation, and intelligent file organization substantially boost sonographer productivity by reducing administrative burden while preserving human control over clinical content. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-powered EHR tools, voice transcription, and automated report drafting meaningfully speed up records maintenance, letting sonographers focus more on clinical tasks while AI handles documentation formatting and retrieval. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Most of the task involves structured data entry, filing, and document management, which modern AI and workflow systems can handle efficiently. However, interpretation of complex clinical context and some medical judgment around what to record may still require human oversight, preventing a full 5-rating. |
| Task automatability | claude-sonnet-5 | 3/5 | Much of the record-keeping, transcription, and organization of correspondence and reports can be automated with EHR integration and AI text tools, but linking sonographs, interpretations, and quality assurance data still requires human verification and judgment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Healthcare records are subject to HIPAA and institutional compliance requirements that necessitate human oversight and audit trails. While not a complete bar, these regulatory and liability concerns create meaningful friction and require system validation before full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Patient record accuracy and regulatory compliance (HIPAA, quality assurance standards) impose oversight requirements, though the task itself is largely administrative and not exclusively reserved for licensed personnel to perform every step. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-driven EHR tools and document management systems cost substantially less than paying a sonographer's loaded wage ($50–70k+ annually) to perform clerical record-keeping. Integration and oversight add some cost, but the ratio remains highly favorable for automation. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-assisted documentation tools reduce clerical time but still require licensed staff oversight and integration costs, making the net cost savings moderate rather than dramatic. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Electronic health record (EHR) systems with AI-assisted documentation, OCR, and automated data extraction are deployed in many healthcare facilities today. Clinical note generation and image metadata management are production-ready, though integration with specific institutional systems varies. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | EHR and PACS systems with AI-assisted documentation are deployed in many hospitals, but full automation of maintaining diverse record types (biopsy, post-op reports, correspondence) with accuracy is not yet standard practice. |
Process and code film from procedures and complete appropriate documentation.
56CI 45–67 · exposure 58 · augmentation 75 · importance 4.5/5 · click for rater detail
Process and code film from procedures and complete appropriate documentation.
56| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare is moderately digitized and has active adoption of AI workflow solutions, but sonography practices show uneven adoption. Large health systems deploy automated documentation more readily, while independent clinics and smaller practices lag, resulting in middling sector-wide velocity. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative AI adoption is growing but healthcare IT systems are often slow to modernize, with documentation AI in pilot or partial-rollout stages rather than fully deployed at scale. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI substantially assists sonographers by auto-populating documentation, suggesting procedure codes, and organizing film archives, freeing human time for patient interaction and clinical review. The sonographer remains in the loop for verification and complex cases, raising overall productivity significantly. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully speed up coding suggestions, auto-populate documentation fields, and flag inconsistencies, improving sonographer efficiency while they retain final review responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can reliably extract, categorize, and code ultrasound/sonography images and automatically populate structured documentation fields, achieving significant time savings. The task involves routine procedural coding and documentation that maps well to machine-learnable patterns, though some edge cases and complex clinical nuances may still require human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | Coding and documentation of imaging procedures involves structured data entry and standardized terminology that AI can partially handle, but linking codes to specific findings and images still requires human verification. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Moderate barriers exist: regulatory compliance (HIPAA, medical record accuracy standards) and liability concerns around coding accuracy require oversight; many facilities require clinician sign-off on documentation. However, the task itself does not require a licensed sonographer to personally perform the coding and filing, reducing hard legal barriers. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Documentation must meet regulatory and billing compliance standards, and errors carry liability risk, creating moderate friction even though the task itself is not inherently requiring hands-on patient contact. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Automated image processing and documentation coding costs (inference + integration) are typically an order of magnitude lower than the loaded wage of a sonographer performing these clerical tasks, especially at scale across high-volume clinics and hospitals. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-assisted coding tools can reduce clerical time but still require licensed staff oversight and integration with PACS/EHR systems, keeping costs comparable to human labor in many settings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed DICOM-compliant AI systems and radiology workflow software already perform automated image tagging, filing, and structured documentation in production healthcare settings. Products like those from major PACS vendors and specialized sonography AI vendors reliably handle image processing and coding, though integration with legacy systems varies. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some medical coding/documentation assistance tools exist but are not widely deployed specifically for sonography image processing and coding workflows with high reliability. |
Obtain and record accurate patient history, including prior test results or information from physical examinations.
37CI 25–50 · exposure 38 · augmentation 63 · importance 4.6/5 · click for rater detail
Obtain and record accurate patient history, including prior test results or information from physical examinations.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare sectors show moderate adoption of AI-assisted documentation, but autonomous history-taking remains rare in production. Sonography practices continue to rely on direct human-patient interaction for history collection, with limited evidence of rapid AI displacement in this specific task. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare is a middling-adoption sector for AI documentation tools—pilots and partial deployment (ambient scribes) are growing but not yet the default across sonography practices. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by summarizing prior test results, flagging relevant historical patterns, and auto-populating structured fields from existing records, which reduces data entry burden and helps sonographers prepare for patient interviews more efficiently. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scribing and transcription tools meaningfully speed up history capture and documentation, letting sonographers focus more on patient interaction and exam quality while AI drafts notes. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can extract and organize structured data from medical records and prior test results, but obtaining accurate patient history requires real-time clinical judgment, nuanced questioning, and the ability to probe contradictions or vague responses—capabilities current AI systems lack reliably. Significant human oversight would be needed to verify completeness and accuracy. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can transcribe, summarize, and extract structured history from patient interviews or records via ambient documentation tools, but eliciting nuanced clinical history and verifying accuracy still typically requires clinician judgment and interaction. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Patient history taking in clinical settings carries legal, liability, and regulatory constraints. Sonographers are expected to document information with accountability; AI systems cannot independently sign off on patient history accuracy without licensed human verification, creating a hard requirement for human review and sign-off. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensure requirement for taking history itself, but clinical liability, EHR integration requirements, and patient trust/interaction norms create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI tools for medical record extraction are relatively inexpensive, but when integrated into clinical workflows with necessary quality oversight and verification by human sonographers, the all-in cost approaches parity with direct human labor. Significant data validation work remains manual. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI scribing tools reduce documentation time and cost, but licensing, integration with EHRs, and required human review keep costs roughly comparable to trained staff for this narrow task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can parse existing medical records and extract key data points, no deployed product reliably conducts independent patient interviews or validates historical information in clinical settings. Products exist for clinical documentation support but do not autonomously obtain and synthesize patient history at the level sonographers perform. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Ambient clinical documentation and voice-to-text tools (e.g., DAX, Nuance) are deployed in some healthcare settings to capture history, but sonography-specific workflows and full end-to-end accuracy verification are not yet mature or universal. |
Provide sonogram and oral or written summary of technical findings to physician for use in medical diagnosis.
28CI 25–30 · exposure 30 · augmentation 63 · importance 4.8/5 · click for rater detail
Provide sonogram and oral or written summary of technical findings to physician for use in medical diagnosis.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI sonography tools remains limited and concentrated in large academic medical centers and specialized practices. Most sonography departments operate with human technologists; uptake of fully autonomous or high-automation systems is slow due to regulatory friction, liability concerns, and clinical conservatism in diagnostic imaging. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is a highly regulated, moderately digitized sector where AI adoption for diagnostic tasks is proceeding cautiously through FDA-cleared point tools rather than broad autonomous deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI augmentation is emerging in the form of real-time measurement assistance, anatomical landmark detection, and abnormality flagging during scanning, which can guide sonographers and improve efficiency and consistency. However, the human sonographer remains central to probe control, patient interaction, and quality assessment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-based image quality enhancement, automated measurements, and draft report generation meaningfully speed up sonographers' workflow and improve consistency while the sonographer remains responsible for final findings. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI can assist in image analysis and flagging abnormalities, but the task requires integration of patient context, clinical judgment in interpretation, and direct physician communication. End-to-end automation with 50% time savings at equal quality is not demonstrated; human sonographers remain necessary for probe positioning, real-time decision-making, and quality assurance. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can assist with image analysis and draft summaries, but capturing quality diagnostic images and synthesizing clinical findings from real-time scanning still requires human judgment and physical scanning skill not automatable end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Substantial barriers exist: medical licensure and liability requirements mean a licensed clinician must oversee and sign off on findings; regulatory frameworks (FDA, clinical standards) govern autonomous ultrasound devices; patient care standards favor human contact and real-time probe adjustment, and reimbursement typically requires human credentialing. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Medical imaging diagnosis requires licensed practitioner sign-off and is subject to strict regulatory and liability requirements, creating strong barriers to full automation of the diagnostic communication step. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI analysis tools require licensing, integration with ultrasound systems, and significant oversight by sonographers or radiologists. When including infrastructure, training, and the human review still needed, the all-in cost remains comparable to or higher than hiring sonography labor in many settings. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Sonographer time and skilled scanning cannot be replaced by cheap inference alone; AI tools add cost as adjuncts requiring integration with ultrasound hardware and human oversight, so all-in cost is not clearly cheaper than the sonographer's labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | AI products exist for sonogram analysis (e.g., cardiac, obstetric screening tools) with reasonable accuracy on well-defined findings, but error rates and narrow applicability (specific organs/pathologies) limit reliability at scale. Deployment is primarily as a second-reader or aid, not as autonomous end-to-end performance. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI-assisted image interpretation tools exist for specific applications (e.g., cardiac ejection fraction, obstetric measurements) but no deployed product autonomously performs full sonogram interpretation and reporting at scale in routine clinical practice. |
Record and store suitable images, using camera unit connected to the ultrasound equipment.
27CI 21–34 · exposure 33 · augmentation 63 · importance 4.6/5 · click for rater detail
Record and store suitable images, using camera unit connected to the ultrasound equipment.
27| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare adoption of AI for core diagnostic imaging tasks remains cautious and heavily regulated. While image analysis tools are emerging, automation of the image-capture process itself is still nascent, with most facilities relying on traditional sonographer workflows. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging adopts AI slowly due to regulatory approval processes, liability concerns, and the hands-on nature of ultrasound acquisition, despite AI tools for image analysis expanding. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist sonographers through real-time image quality feedback, automated archiving, and guidance on anatomical landmarks, meaningfully improving workflow efficiency. However, the augmentation is partial—the sonographer retains control of probe manipulation and clinical judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted image quality feedback and automatic capture triggers on modern ultrasound systems help sonographers ensure suitable images are stored more efficiently, meaningfully aiding the task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in image capture and quality assessment, the task requires real-time judgment about positioning, probe angle, and patient movement—decisions that currently demand human supervision. Full end-to-end automation without sonographer oversight does not meet the 50% time-saving bar. |
| Task automatability | claude-sonnet-5 | 3/5 | The act of capturing and storing images is largely mechanical, but selecting which images are 'suitable' requires clinical judgment about diagnostic quality, limiting full automation today.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Ultrasound imaging requires licensure (Registered Diagnostic Medical Sonographer credentials in most jurisdictions) and clinical accountability for diagnostic image quality. Legal and liability frameworks mandate human credentialing and responsibility for the images recorded. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Diagnostic imaging is part of a licensed clinical procedure requiring a credentialed sonographer to operate equipment and ensure image adequacy for physician interpretation, creating strong regulatory and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The equipment cost and integration overhead for any autonomous ultrasound system would far exceed the incremental labor cost of a human sonographer managing image capture and storage, making AI substantially more expensive all-in. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | The equipment itself is a sunk capital cost already used by sonographers; there's no separate AI system replacing this narrow sub-task at lower cost since the sonographer must still perform the scan. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product autonomously performs ultrasound image capture and storage at diagnostic quality. Automated quality-check systems exist in research and limited pilots, but reliable end-to-end image acquisition remains dependent on trained sonographer operation. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | While ultrasound machines have automated image capture and storage workflows, no deployed product autonomously judges and selects diagnostically suitable frames without a sonographer's real-time control of the probe and framing. |
Decide which images to include, looking for differences between healthy and pathological areas.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.8/5 · click for rater detail
Decide which images to include, looking for differences between healthy and pathological areas.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI-assisted image analysis in radiology/sonography remains in pilot and early deployment phases; while larger institutions experiment with AI flagging systems, displacement of the image-selection task itself is minimal because human oversight remains mandatory and integrated workflows are still nascent. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging adoption of AI is growing but remains in pilot/narrow-application stages for this specific real-time image curation task, well behind fast-adopting sectors like finance. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist by automatically flagging suspicious regions and highlighting potential abnormalities, which can improve sonographer efficiency in review and reduce fatigue; however, the human sonographer must still make the final clinical judgment on inclusion and significance. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-assisted detection tools can highlight potential abnormalities or suggest optimal image frames, helping sonographers focus attention, though the human retains full decision authority. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI can flag abnormalities in sonographic images with some accuracy, but requires significant human validation and cannot reliably perform the nuanced judgment of clinical relevance—deciding which images to include demands understanding context, clinical history, and subtle pathological markers that AI systems today cannot handle end-to-end at required quality thresholds. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires real-time image judgment during live scanning integrated with probe manipulation and patient interaction; current AI can flag abnormalities but cannot fully replace the sonographer's real-time image selection workflow.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory barriers are substantial: in most jurisdictions, a licensed sonographer or radiologist must review and clinically validate selected images; liability and error-cost asymmetry (missing pathology can be fatal) strongly incentivize human accountability; and diagnostic imaging is covered by FDA/regulatory bodies that generally require credentialed professionals in the loop. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Diagnostic image interpretation errors carry high liability, and licensed sonographers/radiologists are typically required to make or verify these judgments, creating strong regulatory and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference is cheap, but integration, retraining on institution-specific protocols, and mandatory human expert oversight for quality assurance mean total cost per task approaches or exceeds the loaded wage of a sonographer, particularly when liability and malpractice risk are factored in. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Specialized diagnostic imaging AI systems require significant integration, validation, and oversight costs that are not yet cheaper than sonographer time for this specific judgment task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Computer vision models exist for detecting specific pathologies in ultrasound, but deployed products operate with material error rates, narrow scope (trained on specific organs/conditions), and require expert radiologist oversight—no mature end-to-end system reliably performs this full task in production without human sign-off. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI image analysis tools exist for detecting certain pathologies (e.g., cardiac, obstetric) but are narrow in scope and not deployed as autonomous image-selection systems in routine practice. |
Load and unload film cassettes used to record images from procedures.
21CI 15–26 · exposure 8 · augmentation 0 · importance 3.4/5 · click for rater detail
Load and unload film cassettes used to record images from procedures.
21| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare and imaging departments are generally slow to adopt robotics for routine manual tasks; this task remains human-performed in standard practice with minimal automation penetration. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging has been transitioning to digital systems for years, reducing relevance of film cassettes, but adoption of robotics/AI for physical equipment handling in clinical procedure rooms remains slow and rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | There is no meaningful AI assistance possible for a simple, rapid physical handling task that requires only dexterity and doesn't involve decision-making or information processing. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI provides no meaningful assistance for the physical act of loading or unloading film cassettes; this is a manual mechanical task outside AI's current functional domain. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task is purely physical—loading and unloading film cassettes into imaging equipment. Current AI systems cannot perform embodied manipulation tasks reliably in unstructured clinical environments. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a simple physical manipulation task, but it requires hands-on presence with equipment and film; no general-purpose AI system performs this physical action, and most modern sonography has moved to digital capture, making the task itself increasingly obsolete rather than automatable by AI.rating. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While there are no strict licensing barriers to automation, clinical environments have preference for human presence, equipment compatibility concerns, and organizational resistance to replacing routine manual tasks. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing barrier specifically protects this narrow subtask, but it requires physical presence in a clinical environment during a procedure, creating moderate organizational/physical friction against remote automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Humanoid robots or purpose-built automation for this task would cost far more than the technician time saved, especially given the low-skill, high-frequency nature of loading/unloading cassettes. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI system that performs this physical loading/unloading task, so no cost comparison favors AI; a human must physically handle the cassette. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product performs this physical handling task in production healthcare settings. Robotics exist for some manipulation but are not integrated into sonography workflows at scale. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product loads/unloads physical film cassettes; this is a manual physical task, not a cognitive or software task, and remains outside current AI product scope entirely. |
Determine whether scope of exam should be extended, based on findings.
16CI 6–25 · exposure 13 · augmentation 50 · importance 4.7/5 · click for rater detail
Determine whether scope of exam should be extended, based on findings.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI-assisted sonography exists primarily in research and pilot projects; production deployment of AI for clinical decision-making in ultrasound scope determination remains limited, with most radiology and sonography practices relying on human expertise. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is adopting AI for detection assistance but decision-authority tasks like this remain largely untouched by production AI systems due to regulatory and clinical workflow constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist sonographers by flagging potential abnormalities that might be missed, prompting consideration of extended views, but the final decision requires human clinical judgment and responsibility, making this a useful but bounded assistant role. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based image analysis and abnormality detection can flag potential findings that prompt the sonographer to consider extending the exam, offering useful decision support without replacing judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist in identifying abnormal findings that might warrant expanded imaging, but the decision to extend scope requires clinical judgment integrating patient history, contraindications, and real-time clinical reasoning that current systems cannot reliably perform end-to-end with consistent quality. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires real-time clinical judgment integrating patient history, prior images, and evolving findings during a live exam to decide on scope changes—current AI cannot autonomously make or execute such judgment calls in a clinical setting. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Scope extension decisions carry liability risk and regulatory oversight (FDA imaging device rules, scope-of-practice laws), and the sonographer or supervising radiologist bears legal responsibility for whether imaging is adequate—creating hard barriers to full AI autonomy. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a clinical judgment task tied to scope-of-practice regulations and liability; only licensed sonographers/physicians can legally determine exam scope changes. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI image analysis tools require significant oversight and integration costs, and the loaded cost of reliable AI deployment for clinical decision-support remains comparable to or exceeds the incremental labor cost of an experienced sonographer making this judgment. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Even if AI could flag anomalies suggesting extended scope, human oversight and liability requirements mean the AI cannot replace the decision-maker, so cost savings are limited to minor assistive gains rather than full task substitution. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can detect certain abnormalities in sonographic images with moderate accuracy, no deployed product reliably makes scope-extension decisions in clinical production; this remains primarily a sonographer/physician judgment task with limited AI automation in real workflows. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently decides to extend sonographic exam scope; AI image-analysis tools exist but decision-making authority remains with the sonographer/physician. |
Operate ultrasound equipment to produce and record images of the motion, shape, and composition of blood, organs, tissues, or bodily masses, such as fluid accumulations.
14CI 7–21 · exposure 13 · augmentation 50 · importance 4.8/5 · click for rater detail
Operate ultrasound equipment to produce and record images of the motion, shape, and composition of blood, organs, tissues, or bodily masses, such as fluid accumulations.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI in ultrasound is nascent and limited mainly to image-interpretation aids in academic and some hospital settings. The physical operation and image acquisition step itself remains performed by humans in virtually all clinical environments, with minimal AI displacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is a physical, highly regulated environment with slow uptake of autonomous scanning technology despite AI use in image analysis support tools. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist sonographers by flagging potential quality issues, suggesting anatomical landmarks, or providing real-time measurement aids, but augmentation is modest because the operator must still master probe handling, patient communication, and live decision-making during the scan. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-assisted image optimization, automated measurements, and quality-guidance tools already help sonographers during scanning, improving efficiency without replacing the physical operation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with image interpretation and quality checks, the core task—physical operation of ultrasound equipment, patient positioning, real-time transducer manipulation, and hands-on assessment of anatomy—requires human presence and dexterity. AI cannot yet autonomously perform the tactile, real-time adjustments needed to acquire diagnostic images. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical task requiring manual probe manipulation, real-time adjustment based on patient anatomy, and tactile feedback that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory bodies (FDA, clinical standards) require licensed sonographers to perform and take responsibility for ultrasound acquisition. Medical liability and credentialing requirements create strong legal barriers: only credentialed humans can sign off on diagnostic imaging protocols. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct patient contact, physical skill, and clinical certification requirements create strong barriers, though not full statutory sign-off requirements comparable to physician diagnosis. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of ultrasound equipment operation—including hardware maintenance, trained technician oversight, and integration of any AI assistants—remains substantially cheaper when performed by a human sonographer than attempting to automate or replace this task today. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-only substitute for the physical scanning process, so cost comparison favors the human sonographer entirely for the core hands-on task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI products exist for post-acquisition image analysis and quality scoring, but no deployed system can autonomously operate ultrasound hardware, position patients, or make real-time probe adjustments. Clinical deployment remains limited to human-in-the-loop interpretation support rather than end-to-end task performance. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously operates ultrasound transducers on patients; robotic ultrasound systems remain research or narrow pilot stage, not production-scale practice. |
Clean, check, and maintain sonographic equipment, submitting maintenance requests or performing minor repairs as necessary.
14CI 10–19 · exposure 8 · augmentation 25 · importance 4.3/5 · click for rater detail
Clean, check, and maintain sonographic equipment, submitting maintenance requests or performing minor repairs as necessary.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare facilities have not adopted autonomous robotic systems for equipment maintenance at meaningful scale; adoption remains in pilot/research phase. Physical equipment care in clinical settings lags far behind information-worker automation. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Physical equipment maintenance in clinical settings sees minimal AI adoption; this is a low-digitization, hands-on task with no robotics deployment at scale. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by monitoring equipment logs, recommending maintenance schedules, or highlighting anomalies detected in performance data, but the human sonographer must still perform the physical work. Augmentation is limited to decision support rather than transforming the hands-on portion. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could potentially assist with predictive maintenance alerts or digital logging of issues, but offers little direct help with the physical cleaning and repair actions themselves. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI systems could potentially document equipment status or flag obvious issues via image analysis, the physical inspection, cleaning, and hands-on repair/maintenance work requires embodied robotics that are not reliably deployed today. Only portions (e.g., automated log review) are feasible. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical, hands-on task involving cleaning equipment, visual/functional checks, and manual minor repairs, none of which current AI systems can perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Equipment maintenance in healthcare faces moderate friction: regulatory compliance requirements around device upkeep, liability for improper repairs affecting patient safety, and organizational preference for qualified staff sign-off. These provide some protection but are not hard legal mandates requiring a licensed human. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not licensed work per se, equipment maintenance on medical devices involves safety protocols, manufacturer requirements, and facility policies that create moderate procedural friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Robotic systems capable of equipment maintenance and repair are expensive to deploy and integrate, while a sonographer's routine maintenance takes modest time and incurs no additional cost beyond their wage. AI solutions remain far costlier. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical maintenance work, so AI cost is not comparable—human labor remains the only viable option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system reliably performs end-to-end physical equipment maintenance, inspection, and minor repair work in hospital settings today. This task requires physical manipulation and mechanical troubleshooting beyond current deployed automation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product cleans or physically repairs medical imaging equipment; this remains entirely a human physical maintenance task. |
Observe screen during scan to ensure that image produced is satisfactory for diagnostic purposes, making adjustments to equipment as required.
14CI 3–25 · exposure 13 · augmentation 63 · importance 4.9/5 · click for rater detail
Observe screen during scan to ensure that image produced is satisfactory for diagnostic purposes, making adjustments to equipment as required.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ultrasound departments remain human-centric with slow technology diffusion; most facilities still rely on sonographer expertise rather than algorithmic quality control, and pilot adoption of such systems is limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is adopting AI for image analysis and quality-assist tools, but physical scanning tasks in radiology/sonography departments show slow deployment of scanning automation, reflecting the sector's cautious, regulated pace. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can effectively assist sonographers by providing real-time feedback on image quality, highlighting potential artifacts or out-of-protocol imaging, and suggesting positioning adjustments, materially improving workflow efficiency while the sonographer retains decision authority. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based image quality feedback and probe-guidance tools can help sonographers identify suboptimal images and refine technique in real time, offering moderate assistance without replacing the sonographer's active scanning role. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can analyze ultrasound image quality and detect some artifacts or inadequacies, but real-time equipment adjustment requires nuanced judgment about probe positioning, pressure, and angle that is difficult to standardize across anatomical variations. Current systems lack the embodied control loop integration needed for end-to-end automation with 50%+ time savings. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires real-time hands-on manipulation of the ultrasound probe, physical patient contact, and continuous visual judgment plus instrument adjustment during live scanning; no off-the-shelf system performs this physical, real-time control loop today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory oversight (FDA approval for diagnostic imaging devices), professional licensing requirements for sonographers, and liability concerns over who is responsible for image adequacy create substantial legal and organizational barriers to autonomous operation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Sonography requires licensed/certified professionals for diagnostic imaging, direct physical patient contact, and legal responsibility for image quality and diagnostic adequacy, creating strong regulatory and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration of AI quality-assurance systems into ultrasound machines requires significant hardware and software engineering; the operational cost remains comparable to or higher than the direct labor of a sonographer performing manual quality checks. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task alone, so cost comparison favors the human sonographer who must be present regardless; any AI would only add cost as an assistive layer, not a replacement. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Image quality assessment tools exist in research and some commercial ultrasound platforms, but they operate as assistive overlays rather than autonomous decision-makers. No deployed product reliably performs real-time quality control and automated equipment adjustment without human intervention in production settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | There are no deployed products that autonomously operate the transducer and adjust equipment settings during a live scan; existing AI tools only assist post-hoc image analysis or guidance overlays, not physical control. |
Select appropriate equipment settings and adjust patient positions to obtain the best sites and angles.
12CI 3–21 · exposure 13 · augmentation 50 · importance 4.8/5 · click for rater detail
Select appropriate equipment settings and adjust patient positions to obtain the best sites and angles.
12| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare is adopting AI for image interpretation and decision support, but physical patient-facing examination tasks in clinical settings remain largely manual; automation of positioning and probe adjustment has not reached production adoption. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is a moderately digitized sector but hands-on diagnostic procedures see slow AI adoption compared to administrative or interpretive tasks in the field. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI systems can augment sonographers significantly by recommending optimal angles, suggesting equipment settings based on anatomy, and highlighting anatomy landmarks—freeing the sonographer to focus on probe manipulation and patient comfort. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Some AI-assisted probe guidance and image optimization tools exist to suggest adjustments, but they provide limited assistance to the core physical positioning task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in image analysis and suggest optimal angles based on anatomy, the task requires real-time physical manipulation of ultrasound probes and patient positioning—both inherently manual operations that cannot be automated end-to-end with 50% time savings by current AI systems. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires real-time physical manipulation of an ultrasound probe, hands-on patient positioning, and continuous adjustment based on live tissue feedback—tasks requiring physical presence that current AI cannot perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and regulatory barriers exist: ultrasound involves direct patient contact and medical judgment; only licensed sonographers can legally perform the examination and adjust patient positioning for safety and diagnostic accuracy. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Requires licensed/certified sonographer judgment, direct physical patient contact, and clinical accountability for image quality and patient safety, creating strong barriers to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Automating this task would require robotics and specialized hardware integration far exceeding the cost of a human sonographer's labor, making AI implementation orders of magnitude more expensive. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical task, so cost comparison favors the human sonographer entirely; robotic ultrasound systems remain experimental and expensive. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI image analysis tools exist for post-acquisition optimization, but no deployed product reliably performs the full task of selecting equipment settings and adjusting patient positions autonomously; this remains research-stage or narrow-scope (e.g., guidance overlays only). |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously positions patients or physically manipulates transducers; this remains a manual clinical skill performed by trained sonographers. |
Supervise or train students or other medical sonographers.
4CI 0–7 · exposure 0 · augmentation 38 · importance 4.0/5 · click for rater detail
Supervise or train students or other medical sonographers.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare training and supervision remain dominated by human practitioners due to regulatory requirements, accreditation standards, and the centrality of human mentorship in clinical competency development. Adoption of AI for this specific task is negligible. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare training programs are cautious adopters of AI for clinical education, with pilots in simulation but little production use for direct supervision. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with scheduling, documentation of training sessions, or providing supplementary learning materials, but these are peripheral to the core supervisory relationship. The primary task—evaluating competence and guiding professional development—offers limited augmentation potential. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based simulators, image quality feedback tools, and e-learning modules can supplement training materials and self-assessment, aiding but not replacing human supervision. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Supervising and training medical professionals requires real-time interaction, personalized feedback, assessment of competence, and adaptive pedagogical judgment—capabilities that current AI systems cannot perform end-to-end. This task is fundamentally relational and requires human judgment about skill gaps and individual learning needs. |
| Task automatability | claude-sonnet-5 | 1/5 | Supervising and training sonography students requires hands-on demonstration of probe technique, real-time correction of scanning posture, and clinical judgment feedback that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Medical training and credentialing are heavily regulated; supervisors and trainers must typically hold specific licenses and credentials. Liability and patient safety concerns create strong legal and organizational barriers to AI performing this supervisory function. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Clinical training and supervision typically require credentialed, experienced sonographers per accreditation and hospital policy, creating strong institutional and licensing-adjacent barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of deploying an AI system capable of reliable supervision, including oversight and liability management, would far exceed the loaded wage of an experienced sonographer providing this training in typical healthcare settings. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this supervisory/training role, so cost comparison favors the human trainer entirely. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs supervisory or training roles for medical technicians in production settings. While AI can provide educational content or assist with documentation, actual supervision and competency assessment remain human-dependent. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product supervises or trains sonographers autonomously; educational AI tools exist only as supplementary aids, not replacements for clinical preceptors. |
Prepare patient for exam by explaining procedure, transferring patient to ultrasound table, scrubbing skin and applying gel, and positioning patient properly.
3CI 0–5 · exposure 0 · augmentation 0 · importance 4.7/5 · click for rater detail
Prepare patient for exam by explaining procedure, transferring patient to ultrasound table, scrubbing skin and applying gel, and positioning patient properly.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This task occurs in healthcare settings with strict regulatory oversight, small adoption of physical automation, and strong requirements for human presence. Healthcare remains a laggard sector for hands-on clinical task automation. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare physical patient-handling tasks show minimal AI/robotic adoption industry-wide; this is a laggard area due to safety, cost, and physical complexity. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | AI systems offer no meaningful assistance in explaining procedures to patients, physically transferring them, or positioning them on equipment—all core components requiring direct human judgment and physicality. |
| Augmentation potential | claude-sonnet-5 | 1/5 | Current AI tools offer negligible assistance for the physical aspects of patient transfer, skin prep, and positioning, though verbal explanation scripts could marginally help communication. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical manipulation of patients (transferring, positioning, applying gel), direct patient communication, and judgment about comfort—capabilities that current robots and AI systems cannot perform reliably in clinical settings today. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical patient handling (transferring, positioning), physical contact (scrubbing skin, applying gel), and in-person communication, none of which current AI systems can perform without robotic embodiment far beyond deployed capability. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Patient contact is mandatory and legally required; a licensed healthcare professional must directly interact with the patient for consent, communication, and safe handling. Liability and patient safety regulations create hard barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct physical patient contact and safe transfer/positioning typically require a trained, often licensed healthcare worker due to patient safety and liability concerns, creating strong practical and regulatory barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of deploying a robotic system capable of safe patient handling, combined with required safety redundancy and oversight, far exceeds the loaded wage of a sonographer's preparation work. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for the physical labor involved, so any hypothetical robotic solution would be far more expensive than a sonographer performing this routine physical task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed commercial system can autonomously perform patient transfer, skin preparation, and positioning with the safety and legal requirements necessary in medical practice. This remains entirely human-performed in real clinical settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI or robotic product performs patient transfer, skin prep, or physical positioning in clinical sonography settings; this remains purely manual work. |
Coordinate work with physicians or other healthcare team members, including providing assistance during invasive procedures.
1CI 0–3 · exposure 0 · augmentation 25 · importance 4.4/5 · click for rater detail
Coordinate work with physicians or other healthcare team members, including providing assistance during invasive procedures.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare adoption of AI for autonomous team coordination and procedural assistance remains minimal; medical procedures require licensed human presence by law and practice standards, preventing displacement of this task. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare's clinical/physical care settings show slow AI adoption for hands-on procedural work, even though administrative and imaging-analysis tasks in the sector see faster uptake. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist with scheduling or documentation coordination, but the core task of real-time procedural teamwork and hands-on assistance offers limited room for AI augmentation while a human sonographer remains responsible. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI may assist with documentation, checklists, or image interpretation support around the procedure, but offers minimal direct assistance for the physical coordination and hands-on assistance itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time coordination, interpersonal communication, and physical presence during invasive medical procedures—capabilities far beyond current AI systems. No AI can autonomously manage team workflows or provide hands-on procedural assistance. |
| Task automatability | claude-sonnet-5 | 1/5 | This is real-time, hands-on physical coordination and assistance during invasive medical procedures, requiring physical presence, manual dexterity, and split-second teamwork that current AI cannot perform. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Medical licensure, liability laws, and informed consent requirements mandate that a credentialed sonographer be physically present and responsible for patient safety during procedures. Regulatory and legal frameworks create hard barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Direct patient care during invasive procedures requires licensed, credentialed personnel physically present, with strict regulatory, liability, and safety requirements precluding automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task involves human coordination and physical procedural assistance that would require human workers regardless; AI has no meaningful cost advantage in replacing this human-centered, synchronous activity. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical/interpersonal task, so cost comparison favors the human by default since AI cannot deliver the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably coordinates medical team workflows or assists during invasive procedures in production environments. This requires human judgment, presence, and accountability that current systems cannot provide. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product provides physical assistance or real-time interprofessional coordination during invasive procedures; this remains entirely outside current AI product capability. |
Observe and care for patients throughout examinations to ensure their safety and comfort.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.8/5 · click for rater detail
Observe and care for patients throughout examinations to ensure their safety and comfort.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare organizations cannot and do not substitute AI for patient safety monitoring; regulatory and professional standards mandate continuous human oversight during diagnostic procedures. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare imaging is a highly regulated, low-digitization-for-physical-care sector with minimal automation of hands-on patient monitoring tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI might provide minor assistance (e.g., alerts if vital signs show anomalies on connected monitors) but the sonographer's core duty—direct observation and personal care—remains fundamentally human-centered and resists meaningful augmentation. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can support documentation or alert systems (e.g., vital sign monitoring), but it offers limited direct assistance to the core act of physically observing and comforting patients. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Observing and caring for patient safety and comfort requires real-time physical presence, emotional intelligence, and responsive adaptation to patient distress—capabilities current AI systems fundamentally lack. No automation pathway exists for this inherently human-contact task. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires direct physical presence, hands-on positioning, real-time comfort assessment, and emergency response capability that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Medical licensure and professional liability standards legally require a qualified human sonographer to remain present and responsible for patient safety and comfort throughout the examination. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Patient safety, physical contact, and clinical licensure requirements make this a hard barrier requiring a credentialed human present at all times. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems cannot physically be present with patients or provide comfort care, making cost comparison inapplicable; the task remains firmly in human domain. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical caregiving task, so no meaningful cost comparison favors AI; a human sonographer is required regardless of cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product can replace a human sonographer's duty to monitor a patient's physical and emotional state during an examination. This task requires licensed human presence and accountability. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product monitors and physically cares for patients during sonographic exams; this remains entirely a human clinical function. |
Perform legal and ethical duties, including preparing safety or accident reports, obtaining written consent from patient to perform invasive procedures, or reporting symptoms of abuse or neglect.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.0/5 · click for rater detail
Perform legal and ethical duties, including preparing safety or accident reports, obtaining written consent from patient to perform invasive procedures, or reporting symptoms of abuse or neglect.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | There is no meaningful AI adoption in this area because legal and ethical duty cannot be transferred to automation. Regulatory frameworks explicitly require human accountability, making velocity immaterial. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare compliance and legal-ethical documentation processes see minimal AI-driven displacement due to liability, regulation, and the need for human accountability in these specific duties. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist marginally with documentation templates, flagging language patterns suggestive of abuse, or organizing incident report data, but the core task—judgment, consent, and legal responsibility—must remain with the human sonographer. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can help draft or template accident/safety reports and consent documentation, but the core legal-ethical actions—obtaining consent and judgment calls on reporting abuse—still require human execution with limited AI assistance. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires human judgment, legal authority, and ethical responsibility in deciding what constitutes abuse or neglect and obtaining informed consent—areas where AI cannot operate autonomously or legally. Current AI systems cannot make binding ethical or legal determinations or serve as the responsible party for consent documentation. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires legal judgment, patient interaction, mandatory reporting obligations, and consent processes that are inherently human-executed and legally attributed to the licensed practitioner; AI cannot perform these end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and regulatory barriers mandate that a licensed healthcare professional must personally obtain consent, file incident reports with legal standing, and fulfill mandatory reporting obligations. These are non-delegable human responsibilities under healthcare and child protection law. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legal consent, licensure accountability, and mandatory reporting laws for abuse/neglect create hard regulatory and liability barriers requiring a human professional to perform and be accountable for these duties. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | A sonographer must personally perform these duties by law and ethical requirement; there is no cost comparison, as AI cannot substitute for the human responsibility holder. The task cannot be cost-displaced. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot substitute for the legally required human actions (consent, reporting), there is no viable AI cost basis to compare against the human wage for this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs these functions independently; AI cannot obtain valid patient consent, file official safety reports with legal standing, or make mandatory abuse reporting decisions. These require human accountability and legal authority that AI systems do not possess. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously obtains legally valid patient consent or files mandatory abuse/neglect reports on a sonographer's behalf; this remains squarely a human legal and ethical responsibility. |
Perform medical procedures, such as administering oxygen, inserting and removing airways, taking vital signs, or giving emergency treatment, such as first aid or cardiopulmonary resuscitation (CPR).
0CI 0–0 · exposure 0 · augmentation 25 · importance 3.9/5 · click for rater detail
Perform medical procedures, such as administering oxygen, inserting and removing airways, taking vital signs, or giving emergency treatment, such as first aid or cardiopulmonary resuscitation (CPR).
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No meaningful adoption of AI for autonomous execution of these procedures exists in healthcare, nor is it anticipated. The sector remains entirely dependent on licensed human clinicians for these emergency and critical care tasks. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare physical intervention tasks show minimal AI displacement; adoption in this specific hands-on emergency care context is essentially nonexistent. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could offer minimal assistance in reminding clinicians of CPR protocols or vital-sign thresholds, but these tasks are so procedurally standardized and time-critical that decision support offers limited added value. The human clinician's training and muscle memory dominate. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can support decision protocols or monitoring device alerts, but offers little direct assistance during the physical execution of airway insertion or CPR itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires direct physical manipulation of patient bodies (oxygen delivery, airway insertion, vital sign measurement, CPR) and real-time clinical judgment in emergency settings. Current AI systems cannot perform these hands-on interventions; they exist only as software or control systems without embodied agency in medical environments. |
| Task automatability | claude-sonnet-5 | 1/5 | This is hands-on physical medical intervention requiring manual dexterity, real-time judgment, and direct patient contact; no current AI system can physically administer oxygen, insert airways, or perform CPR. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | These medical procedures are legally restricted to licensed healthcare professionals (nurses, physicians, paramedics) and carry high liability. Regulatory bodies (state medical boards, OSHA) and malpractice liability create hard legal barriers; human licensure and direct patient contact are non-negotiable. |
| Adoption barriers | claude-sonnet-5 | 5/5 | These are licensed clinical procedures with strict legal, safety, and liability requirements mandating trained human personnel, especially in emergencies. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI cannot perform these tasks at all today, making cost comparison moot. Any attempt would still require a human clinician to actually execute the physical interventions, so the AI cost is purely supplementary if it exists. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing these physical interventions, so cost comparison is moot—human labor is the only option today. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can autonomously administer oxygen, insert airways, take vital signs via direct patient contact, or perform CPR. These require licensed human clinicians and physical presence; research robots exist but do not perform these tasks reliably in clinical practice at scale. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs these physical emergency medical procedures autonomously; robotics for such tasks remain research-stage at best. |
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.