Nuclear Medicine Technologists
29-2033.00Median wage $101,370/yr17,080 employed (US)Rank #787 of 923 scored · top 85% by substitution
Prepare, administer, and measure radioactive isotopes in therapeutic, diagnostic, and tracer studies using a variety of radioisotope equipment. Prepare stock solutions of radioactive materials and calculate doses to be administered by radiologists. Subject patients to radiation. Execute blood volume, red cell survival, and fat absorption studies following standard laboratory techniques.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution15
Exposure17
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
17 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
0%
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%18
panel mean rating 1.7/5 → substitution pressure 18/100
Technical feasibility todayw 20%15
panel mean rating 1.6/5 → substitution pressure 15/100
Cost vs. human wagew 15%16
panel mean rating 1.6/5 → substitution pressure 16/100
Adoption barriersw 20%inverted — strong barriers lower the score12
panel mean rating 4.5/5 (barrier strength) → substitution pressure 12/100
Sector adoption velocityw 10%14
panel mean rating 1.5/5 → substitution pressure 14/100
Task breakdown (17 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.
Record and process results of procedures.
51CI 39–62 · exposure 58 · augmentation 75 · importance 4.8/5 · click for rater detail
Record and process results of procedures.
51| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare IT adoption of AI for routine recording and processing is progressing slowly; while large academic medical centers pilot AI integration, most community hospitals and independent nuclear medicine facilities rely on legacy RIS/PACS with minimal autonomous processing automation. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare documentation is adopting AI-assisted tools steadily but unevenly, with pilots common in EHR/RIS integration but full automation of results processing still limited in most facilities. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists technologists by automating routine data entry, flagging potential imaging artifacts, organizing results, and suggesting quality issues, allowing technologists to focus on verification and clinical judgment rather than manual transcription and organization. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted transcription, templated reporting, and automated data extraction significantly speed up how technologists record and process procedure results while they remain responsible for accuracy. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate significant portions of result recording and routine processing (data entry, image organization, basic quality checks) but clinical interpretation and decision-making about procedural outcomes typically require human verification, limiting full end-to-end automation to roughly 50% time savings. |
| Task automatability | claude-sonnet-5 | 4/5 | Recording and processing procedure results (dosage logs, imaging metadata, structured reports) is largely data entry and documentation that current AI/automation systems can handle with high time savings, though final verification remains human.4 |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Nuclear medicine result processing involves regulatory requirements (FDA oversight of diagnostic AI, accreditation standards), liability concerns (incorrect results carry clinical risk), and legal/medical necessity that a licensed technologist verify and sign off on recorded findings, creating strong substitution barriers. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Medical record-keeping requires accuracy and regulatory compliance (HIPAA, radiation safety documentation), and technologists are often required to verify and sign off on records, creating moderate barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI integration into nuclear medicine workflows carries substantial overhead (specialized medical AI licensing, integration with existing hospital systems, validation requirements), making per-task costs comparable to or slightly exceeding technologist labor for this specific workflow. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated logging and templated report generation software is inexpensive per use compared to technologist time spent on manual documentation, though integration and oversight costs remain. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Medical imaging analysis and data processing products exist in production (RIS/PACS systems with some AI modules), but reliable fully autonomous processing of nuclear medicine results with sufficient clinical accuracy remains inconsistent; most deployments require human oversight. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | RIS/PACS systems and speech-to-text/structured documentation tools are deployed in radiology/nuclear medicine departments today, but full automated processing of results with no technologist review is not yet standard practice. |
Process cardiac function studies, using computer.
34CI 25–43 · exposure 38 · augmentation 63 · importance 4.8/5 · click for rater detail
Process cardiac function studies, using computer.
34| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare and nuclear medicine departments adopt AI slowly relative to other sectors; pilots are emerging but production deployment of end-to-end automation is rare, with strong preference for human technologist oversight of clinical image processing. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging adoption of AI tools is progressing but unevenly, with hospital procurement, regulatory approval (FDA clearance), and integration cycles slowing broad deployment compared to faster-moving digital sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist with automated segmentation, quantification of cardiac metrics, and flagging abnormalities, allowing technologists to work faster and more accurately while remaining responsible for final validation and interpretation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Computer-aided processing software already significantly speeds up quantitative analysis of cardiac studies (e.g., automated gating, volume calculations), meaningfully boosting technologist productivity while they remain responsible for oversight and final quality checks. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While image processing and data analysis components can be partially automated, cardiac function studies require specialized interpretation of nuclear medicine scans that depends on clinical context, patient history, and radiologist review. Current AI cannot reliably handle the full workflow end-to-end without significant human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can process much of the quantitative image analysis and computation for cardiac function studies, but data acquisition setup, quality control, and clinical interpretation oversight still require human involvement, limiting the full end-to-end automation gain. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Nuclear medicine is a regulated medical domain where licensing requirements, liability concerns, and radiologist/physician oversight mandate human involvement in interpretation and sign-off, creating strong legal and organizational barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Nuclear medicine technologists are licensed professionals, and processing diagnostic cardiac studies typically requires certified personnel and physician sign-off due to patient safety and regulatory requirements, creating substantial barriers to full substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Specialized nuclear medicine software, infrastructure, and required human oversight (technologist and physician review) mean total deployment cost remains comparable to or potentially higher than a trained technologist's loaded wage. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Automated processing software is already licensed and integrated into existing workflows, so incremental AI cost is modest, but it doesn't eliminate the technologist's time for setup, QC, and troubleshooting, keeping costs roughly comparable to current practice. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI-assisted image segmentation and analysis tools exist in research and some clinical settings, but production systems for complete cardiac function study processing are not yet widely deployed at scale. Material error rates and integration challenges limit current real-world reliability. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Software packages with automated quantification (ejection fraction calculation, gating analysis) are deployed in nuclear medicine departments, but they require technologist verification and correction, so reliability is moderate rather than fully autonomous. |
Produce a computer-generated or film image for interpretation by a physician.
28CI 16–39 · exposure 30 · augmentation 63 · importance 4.8/5 · click for rater detail
Produce a computer-generated or film image for interpretation by a physician.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare adoption of imaging AI is cautious and gradual, driven by regulatory hurdles and need for clinical validation; most deployments are narrow (image segmentation, artifact reduction) rather than end-to-end production. Technologist roles remain strongly embedded in accredited imaging facilities. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is a moderately digitized but heavily regulated and physically-grounded sector; AI adoption for image reconstruction is growing but the sector overall lags behind information/finance industries in production-scale AI displacement. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist with image reconstruction, noise reduction, quality assessment, and flagging potential retakes, raising technologist productivity in image post-processing and review. However, the assistance is bounded by the need for human judgment on acquisition parameters and clinical acceptability. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-based image reconstruction, artifact reduction, and enhancement tools substantially improve image quality and speed for technologists and interpreting physicians, meaningfully boosting productivity while humans remain in control. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | While image acquisition hardware is automated, the task requires human oversight of patient positioning, equipment calibration, quality control, and clinical judgment about retakes—functions that current AI cannot perform end-to-end. Image processing itself is only one component of a complex workflow tied to patient safety and regulatory compliance. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can assist in image reconstruction and processing, and modern scanner software already automates much of the imaging pipeline, but positioning patients, calibrating dose, and ensuring correct acquisition parameters still require human technologist judgment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory frameworks (FDA, ACR, NRC) mandate technologist licensure and direct involvement in imaging procedures; patient contact and safety monitoring are legal requirements. Quality assurance and liability for diagnostic image acceptability rest with licensed personnel, creating hard substitution barriers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Radioactive material handling requires certified/licensed technologists under regulatory bodies (e.g., NRC, state licensure), and patient safety concerns around radiopharmaceuticals create strong human-in-the-loop requirements. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Image processing AI is relatively cheap, but the full value of technologist labor includes patient handling, equipment troubleshooting, safety oversight, and regulatory compliance that AI cannot substitute. The technologist's all-in cost substantially outweighs standalone image processing savings. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Imaging equipment and software licenses represent significant capital cost, and a trained technologist is still required for patient handling and quality control, so AI reduces some labor but does not yet undercut overall costs by an order of magnitude. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI can assist with image reconstruction and post-processing, but no deployed product reliably performs the full task of producing diagnostic-quality nuclear medicine images independently. Clinical deployment requires technologist-supervised acquisition, positioning, and quality verification that AI cannot yet replicate at the required standard. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Deployed nuclear medicine and PET/SPECT systems include automated reconstruction algorithms and AI-based denoising/enhancement tools in production, but full end-to-end autonomous image generation without technologist oversight is not standard practice. |
Detect and map radiopharmaceuticals in patients' bodies, using a camera to produce photographic or computer images.
25CI 25–25 · exposure 25 · augmentation 63 · importance 4.9/5 · click for rater detail
Detect and map radiopharmaceuticals in patients' bodies, using a camera to produce photographic or computer images.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare IT adoption of clinical AI remains measured; nuclear medicine facilities are specialized, and most organizations are still in pilot phases for AI-assisted diagnostics rather than autonomous detection and mapping workflows. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging adopts AI mainly for post-acquisition analysis; the physical scanning task itself sees slow adoption due to regulatory and physical infrastructure constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-driven image enhancement, automated anomaly flagging, and real-time quality feedback can significantly boost technologist productivity and diagnostic confidence, transforming throughput and consistency while the technologist retains protocol control and clinical judgment. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with image reconstruction, noise reduction, and quality control during acquisition, improving efficiency, but the technologist remains essential for the physical operation and patient care aspects. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in image analysis and anomaly detection post-acquisition, the live detection and mapping task requires real-time clinical judgment, patient positioning, protocol adjustment, and quality control that current systems cannot fully automate end-to-end with 50% time savings at equal clinical quality. |
| Task automatability | claude-sonnet-5 | 2/5 | The physical positioning of patients, camera operation, and radiopharmaceutical handling require hands-on manipulation of specialized equipment and patient interaction that current AI cannot perform end-to-end.atable image acquisition workflow remains human-operated. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Nuclear medicine technologists must be licensed/certified and follow strict regulatory protocols (FDA, NRC); the patient-contact requirement, radiation safety protocols, and liability concerns around image quality and patient safety create substantial legal and organizational barriers to autonomous automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Handling radioactive materials and positioning patients for imaging is tightly regulated, requiring licensed nuclear medicine technologists to perform the procedure, creating strong legal and safety barriers to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference for image analysis is inexpensive, but the high-end imaging hardware, integration costs, and required clinical oversight mean total cost per task-equivalent remains comparable to or potentially exceeds the loaded wage of a trained technologist. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Gamma cameras and SPECT/PET systems require significant capital and human oversight; AI software add-ons reduce some processing time but don't replace the technologist's labor cost, so all-in cost is still comparable to or above human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI image analysis tools exist for nuclear medicine scans, but deployed products are limited to post-processing and interpretation support rather than live detection-and-mapping; reliable production systems for autonomous real-time imaging protocol execution and quality assurance remain research-adjacent. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI image reconstruction and enhancement tools exist in some scanners, but the actual acquisition, camera positioning, and patient handling are not performed by deployed autonomous AI products today. |
Perform quality control checks on laboratory equipment or cameras.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.7/5 · click for rater detail
Perform quality control checks on laboratory equipment or cameras.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Nuclear medicine facilities are specialized, hospital-based, and operate in a highly regulated environment with slow technology adoption cycles. The sector is small, conservative about equipment changes, and lacks the digital-first culture seen in finance or software development. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is a highly regulated, moderately digitized sector where automation of physical QC tasks lags behind more administrative or diagnostic-support AI applications. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist with routine data logging, visual flagging of anomalies on camera scans, or automated data trending, raising technologist productivity modestly. However, the core judgment of whether equipment passes QC remains with the human expert, making assistance useful but not transformative. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based analytics can help interpret QC data trends and flag equipment drift, offering moderate assistance while the technologist still performs the physical checks. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Quality control checks on nuclear medicine equipment require visual inspection, calibration verification, and interpretation of test results against complex regulatory standards. While some routine measurements could be partially automated, the task involves judgment calls about equipment fitness and safety that demand human expertise and cannot reliably achieve 50% time savings end-to-end with current AI. |
| Task automatability | claude-sonnet-5 | 2/5 | QC checks involve physical calibration, phantom scans, and hands-on equipment manipulation that current AI cannot fully perform, though some data analysis portions could be automated. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Nuclear medicine equipment QC is subject to strict FDA regulations and licensing requirements; technologists are often certified (CNMT) and must personally certify that equipment meets quality standards. This creates a hard regulatory barrier—a licensed human must sign off on QC results—preventing full substitution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Regulatory and safety requirements (e.g., NRC, accreditation bodies) typically mandate qualified technologists to perform and document equipment QC, creating strong compliance barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The AI systems that could assist with parts of QC (vision inspection, data logging) require integration with specialized equipment and domain-expert oversight, making the all-in cost similar to or higher than the technologist's hourly labor when accounting for calibration, validation, and regulatory compliance time. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | The physical setup, phantom handling, and equipment interaction still require a trained technologist, so AI only reduces a small fraction of labor cost, keeping overall cost comparable to human-only workflows. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Computer vision systems exist for basic visual inspection tasks, but nuclear medicine QC involves specialized equipment (gamma cameras, PET/SPECT systems) and calibration procedures where no mainstream commercial product reliably performs the full QC protocol autonomously. Deployed solutions are limited to narrow subsets (e.g., detecting gross physical defects) rather than comprehensive quality validation. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some software tools assist with analyzing QC data or flagging anomalies, but no deployed product independently performs full physical camera/equipment QC in nuclear medicine labs. |
Gather information on patients' illnesses and medical history to guide the choice of diagnostic procedures for therapy.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.6/5 · click for rater detail
Gather information on patients' illnesses and medical history to guide the choice of diagnostic procedures for therapy.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare, despite digitization, adopts AI cautiously in clinical workflows, particularly in regulated domains like nuclear medicine. Adoption remains in pilot phase for most clinical decision support; production use for autonomous procedure selection is rare. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially imaging/nuclear medicine, is a historically slow-adopting sector for AI decision-support beyond narrow image analysis tools. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-powered EMR search, medical history summarization, and clinical reminder systems can assist technologists in organizing and prioritizing patient information. These tools improve efficiency and reduce documentation burden, though the final clinical judgment remains with the human technologist. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by summarizing EHR data and flagging relevant history, saving technologist time, though the core judgment task remains human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task requires synthesizing patient history, understanding clinical context, and making judgment calls about which procedures to recommend—activities that demand medical knowledge and human interpretation. While AI could assist in retrieving and organizing information, the clinical decision to select diagnostic procedures for therapy requires licensed clinical judgment that current systems cannot reliably provide end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | Extracting structured data from records can be partly automated, but synthesizing patient history to guide diagnostic/therapy choices requires clinical judgment and interaction that current AI cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Nuclear Medicine Technologists operate under state licensure and scope-of-practice regulations; procedure selection involves clinical judgment that is legally and professionally expected to remain under qualified human supervision. Liability and malpractice concerns create strong barriers to unsupervised automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Clinical judgment involving patient history and procedure selection typically requires a licensed technologist/physician, with strong liability and regulatory constraints on diagnostic decision-making. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI systems for medical record review and summarization are not yet cheaper than employing a technologist to gather this history, especially when factoring in the need for human oversight and validation of any AI-generated recommendations. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools could cheaply summarize records, but human oversight, liability, and verification needs keep overall cost comparable to or only modestly below technologist labor costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs independent clinical triage and procedure selection. EMR systems can extract patient history, and NLP can summarize medical records, but end-to-end autonomous selection of appropriate nuclear medicine procedures does not exist in production healthcare settings. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some EHR summarization and clinical decision-support tools exist, but no deployed product reliably gathers and integrates patient history to guide nuclear medicine procedure selection in production. |
Calculate, measure, and record radiation dosage or radiopharmaceuticals received, used, and disposed, using computer and following physician's prescription.
23CI 20–25 · exposure 25 · augmentation 75 · importance 4.8/5 · click for rater detail
Calculate, measure, and record radiation dosage or radiopharmaceuticals received, used, and disposed, using computer and following physician's prescription.
23| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare adoption of AI in radiopharmaceutical dosing remains in pilot phase; most nuclear medicine departments continue manual or semi-automated dose verification. Regulatory scrutiny, liability concerns, and the need for licensed sign-off limit rapid automation even in well-resourced hospitals. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging and nuclear medicine are slower-adopting sectors for full automation due to safety-critical, regulated physical handling requirements, though software-assisted calculation tools are commonly used. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered dose-calculation and record-management systems already assist technologists by automating arithmetic, flagging prescription anomalies, and maintaining compliant audit trails. These tools materially raise technologist productivity and safety when integrated into clinical workflows, even though human judgment and authorization remain required. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Computerized dose calculation and record-keeping systems meaningfully speed up and reduce errors in the documentation and calculation portion of this task, while the technologist remains responsible for physical measurement and compliance. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could automate calculation and record-keeping aspects, the task requires verification of physician prescriptions, safety compliance checks, and real-time adjustment based on patient-specific factors. Current systems lack the integrated clinical judgment needed for reliable end-to-end automation at equal quality without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 2/5 | Dose calculation can be assisted by software, but measurement of actual physical radiopharmaceutical activity, calibration checks, and disposal recording require hands-on physical verification that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Nuclear medicine is heavily regulated; radiopharmaceutical administration is typically under strict FDA, NRC, and state radiation protection laws requiring licensed personnel to prescribe, measure, and authorize dosages. A qualified human (technologist or pharmacist) must legally verify and sign off on radiation dosage before patient administration. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This involves radioactive materials regulated by the NRC/Agreement States; only authorized users and certified technologists can legally handle, measure, and record radiopharmaceutical dosages, creating hard regulatory barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Existing dose-calculation and record-management software requires significant integration with clinical systems, training, and ongoing oversight. The total cost of such systems plus required human validation approaches or exceeds the cost of a technologist performing the task directly, especially at smaller facilities. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Software assistance is cheap, but the physical handling, calibration with dose calibrators, and compliance documentation still require a trained, licensed technologist, keeping overall cost comparable to human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product today reliably performs the full task autonomously. While calculation engines and dose-management software exist in medical facilities, they function as decision-support tools requiring human pharmacists or technologists to validate prescriptions, confirm measurements, and authorize dosages before administration. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Dose calculators and radiopharmacy information systems are common, but these are decision-support tools, not autonomous systems performing physical measurement and disposal logging reliably without technologist verification. |
Develop treatment procedures for nuclear medicine treatment programs.
14CI 13–15 · exposure 16 · augmentation 50 · importance 3.4/5 · click for rater detail
Develop treatment procedures for nuclear medicine treatment programs.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare sectors move cautiously on automation of clinical procedure development due to liability, regulatory burden, and the critical importance of human expertise in nuclear medicine—a specialized, licensed field with strong professional gatekeeping and low digitization of procedural development itself. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized radiological/nuclear medicine practice, is a slow-adopting sector for autonomous AI decision-making due to regulatory and safety constraints, though AI is used for imaging analysis support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by summarizing clinical literature, suggesting procedural templates, and highlighting regulatory requirements, helping technologists draft procedures faster. However, the core task of clinical validation and safety assessment must remain human-led, limiting the transformative upside of augmentation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help technologists by summarizing literature, suggesting dosing calculations, or flagging protocol inconsistencies, improving efficiency while humans retain full responsibility for final procedure design. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Developing treatment procedures requires deep medical knowledge, regulatory compliance, and integration with patient-specific clinical protocols. While AI can assist in synthesizing literature and suggesting procedure frameworks, the task demands clinical judgment and accountability that humans must retain; current systems cannot independently create validated treatment procedures meeting the ≥50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Developing treatment procedures requires clinical judgment, patient-specific dosimetry, and regulatory compliance that current AI cannot autonomously produce end-to-end; at best AI can draft protocol templates for human refinement.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Developing nuclear medicine treatment procedures is tightly regulated (FDA, state licensing boards); procedures must be validated, documented, and typically signed off by licensed physicians or qualified nuclear medicine technologists. Legal and regulatory frameworks mandate human professional responsibility for patient safety protocols. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Nuclear medicine treatment procedures involve radioactive materials and patient safety, requiring licensed medical professionals and regulatory bodies (e.g., NRC, state health departments) to approve and sign off on protocols. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI systems could reduce some research and documentation time, but the cost of integrating, validating, and clinically overseeing AI-generated procedures would be substantial relative to specialist technologist time. The high stakes of medical errors mean oversight costs are non-trivial. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Even where AI could assist in drafting portions of protocols, the need for extensive expert review, validation against safety standards, and liability oversight keeps costs comparable to or only marginally below human-only workflows. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably develops nuclear medicine treatment procedures end-to-end in production clinical settings. This requires domain expertise, regulatory alignment, and physician oversight that goes well beyond current generalist or domain-specific AI capabilities in healthcare. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently designs nuclear medicine treatment procedures; this remains a specialized clinical planning task performed by trained technologists and physicians. |
Explain test procedures and safety precautions to patients and provide them with assistance during test procedures.
10CI 0–20 · exposure 8 · augmentation 38 · importance 4.7/5 · click for rater detail
Explain test procedures and safety precautions to patients and provide them with assistance during test procedures.
10| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare adoption of AI in direct patient-facing clinical roles remains limited and heavily regulated. Patient interaction during nuclear medicine procedures is unlikely to be automated given safety, liability, and licensure requirements that bind human technologists to the task. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare imaging is a highly regulated, in-person clinical setting where AI adoption for direct patient interaction and physical assistance remains slow compared to administrative or diagnostic-support uses. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist marginally by providing pre-visit explanatory videos or supplementary written materials, but live patient interaction during procedures requires human judgment and presence, limiting augmentation impact. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help generate patient education materials, answer FAQs, or produce multilingual explanations beforehand, moderately aiding the technologist's communication workload. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires direct human interaction, empathy, and real-time responsiveness to patient concerns and questions. Current AI cannot reliably perform the physical assistance component or adapt explanations to patient anxiety and individual needs in a live setting. |
| Task automatability | claude-sonnet-5 | 2/5 | Explaining procedures and reassuring/assisting patients requires real-time physical presence, personalized communication, and hands-on support during scanning, which current AI cannot perform end-to-end.wide |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong regulatory and liability barriers exist: patient safety requires a licensed human to be physically present, legal liability for inadequate patient preparation or safety protocols rests with the facility, and clinical best practice mandates direct human-patient contact and informed consent. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Handling radioactive materials and patient safety typically requires a licensed nuclear medicine technologist, and hands-on assistance during a medical procedure has strong regulatory and liability requirements for human presence. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI solutions would require extensive human oversight, integration with clinical workflows, and safety compliance. The cost of building and maintaining such a system would exceed the wage of a technologist who already performs the task directly. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI could cheaply generate informational scripts or handouts, but the physical presence and assistance component still requires a paid technologist, keeping overall cost comparable to human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can autonomously explain procedures to patients in person or provide physical assistance during nuclear medicine tests. While chatbots can answer general health questions, they cannot replace the real-time clinical judgment and hands-on support required in this task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product provides in-person patient explanation and physical assistance during nuclear medicine procedures; at most chatbots give pre-visit informational content. |
Position radiation fields, radiation beams, and patient to allow for most effective treatment of patient's disease, using computer.
10CI 0–20 · exposure 13 · augmentation 63 · importance 4.5/5 · click for rater detail
Position radiation fields, radiation beams, and patient to allow for most effective treatment of patient's disease, using computer.
10| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare and radiation oncology remain low-velocity adopters of autonomous AI for safety-critical clinical tasks. Regulatory conservatism, liability concerns, and the need for human accountability keep this task firmly in the human-controlled domain. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially radiation oncology and nuclear medicine, is a highly regulated, slow-adopting sector for autonomous physical task automation despite growing use of planning software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by suggesting optimal positioning based on imaging data or historical treatment plans, reducing manual planning time and improving accuracy. However, the technologist retains decision authority and must validate all recommendations before execution. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Computer-assisted positioning software, image registration, and treatment planning algorithms significantly aid technologists in achieving accurate and efficient field/beam alignment, improving speed and precision while the human remains in control. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Positioning radiation fields and beams for patient treatment requires real-time spatial judgment, adaptation to patient anatomy, and safety-critical decisions that current AI cannot perform autonomously. This task involves physical coordination with imaging feedback and clinical reasoning that exceeds today's AI capabilities. |
| Task automatability | claude-sonnet-5 | 2/5 | Precise patient and beam positioning requires real-time physical manipulation, verification imaging, and clinical judgment about anatomy that current AI cannot perform end-to-end without a human physically executing and verifying the setup.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Radiation therapy is heavily regulated; only licensed technologists and physicians can authorize and execute patient positioning and beam delivery. Legal, liability, and safety-critical oversight requirements create hard barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a licensed clinical task involving direct patient radiation exposure, requiring certified technologists and physician oversight under strict regulatory and safety frameworks, making full automation legally and practically barred. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI systems, integration, validation, and required human oversight for this safety-critical task would far exceed the loaded wage of a nuclear medicine technologist, given liability and regulatory demands. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Specialized imaging/positioning equipment and software have high capital and integration costs, and human oversight remains mandatory, so cost savings versus a trained technologist are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system today reliably performs autonomous radiation field and beam positioning for treatment. While planning software exists, the actual positioning task—adjusting fields based on live patient state and clinical judgment—remains a human responsibility with no production AI alternative. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some treatment planning systems assist with beam geometry optimization, but actual physical positioning of patients and radiation fields is still manually performed and verified by technologists in production settings. |
Maintain and calibrate radioisotope and laboratory equipment.
7CI 0–14 · exposure 8 · augmentation 25 · importance 4.6/5 · click for rater detail
Maintain and calibrate radioisotope and laboratory equipment.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Nuclear medicine is a highly regulated, low-digitization sector with small specialist teams. Adoption of AI-driven automation in this domain remains minimal; facilities continue to rely on certified technologist oversight and documented procedures. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Clinical nuclear medicine and radiology lab operations are a highly regulated, physically-grounded environment with minimal AI-driven automation of equipment maintenance tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide minor assistance through automated alerts for sensor readings or scheduling recommendations, but the core maintenance and calibration work requires human skill, hands-on adjustment, and regulatory sign-off, limiting augmentation value. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with scheduling calibration logs, flagging anomalies in data trends, or automating record-keeping, but it does not meaningfully change the physical calibration workflow itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could assist with monitoring sensor data and flagging calibration drift, the task requires hands-on physical adjustment of specialized equipment and real-time troubleshooting that demands human presence and judgment. Most of the work involves manual intervention and verification that current AI systems cannot perform end-to-end. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires hands-on physical manipulation of radioactive materials and precision instruments, calibration checks, and regulatory documentation that current AI cannot perform end-to-end without a human physically present. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory oversight by the Nuclear Regulatory Commission, state health departments, and accreditation bodies requires documented human certification and accountability for equipment maintenance records. Liability for equipment failure in a nuclear medicine setting creates strong legal and compliance barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Radioisotope handling and equipment calibration are governed by strict NRC/state regulatory and licensing requirements mandating qualified, certified personnel to perform and document these procedures. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems capable of this task do not exist at scale, and any bespoke robotic solution would be extremely expensive compared to the loaded cost of a trained nuclear medicine technologist performing routine maintenance. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing the physical calibration task, so AI cost is effectively irrelevant/infinite relative to human labor for this specific action. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably maintains or physically calibrates radioisotope and laboratory equipment in production. This task requires embodied robotics and specialized domain knowledge that current general-purpose AI systems lack. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously calibrates radioisotope dose calibrators or nuclear medicine lab equipment; this remains a manual, hands-on technologist responsibility. |
Add radioactive substances to biological specimens, such as blood, urine, or feces, to determine therapeutic drug or hormone levels.
4CI 0–9 · exposure 8 · augmentation 25 · importance 4.4/5 · click for rater detail
Add radioactive substances to biological specimens, such as blood, urine, or feces, to determine therapeutic drug or hormone levels.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Nuclear medicine operates in highly regulated healthcare and laboratory settings with stringent compliance requirements and slow technology adoption cycles. The legal and safety constraints make adoption of any automation for this task essentially non-existent in practice. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Nuclear medicine and clinical laboratory work involves highly regulated, physical, low-digitization tasks with minimal AI/robotic adoption for wet-lab radioactive handling. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While AI might assist with record-keeping, data analysis of results, or protocol documentation, the core physical and chemical task of adding radioactive substances to specimens remains entirely human-dependent. Augmentation potential is minimal given the hands-on nature of the work. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with data logging, calculations, or interpreting resulting assay data, but offers little direct assistance in the physical act of adding radioactive substances to specimens. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires precise physical handling of hazardous radioactive materials, biological specimen preparation, and manipulation of laboratory equipment in a highly regulated environment. Current AI systems cannot perform the hands-on laboratory work, handling of biohazardous materials, or radioactive substance administration that this task demands. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a hands-on lab procedure requiring physical handling of radioactive materials and biological specimens; current AI cannot physically perform pipetting, radiolabeling, or sample preparation.:contentReference[oaicite:0]{index=0} |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Nuclear medicine work is heavily regulated by the Nuclear Regulatory Commission (NRC), state agencies, and occupational safety laws. Technologists must hold specific licenses and certifications, and handling of radioactive materials is legally restricted to authorized personnel. No automation can circumvent these hard regulatory and licensing barriers. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Handling radioactive materials and performing diagnostic lab procedures requires licensure, radiation safety certification, and regulatory compliance (NRC/state rules), creating hard legal barriers to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The infrastructure, regulatory compliance, and specialized equipment required for this task, combined with minimal labor substitution potential, make any AI solution far more expensive than paying a trained nuclear medicine technologist to perform the work directly. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical labor and specialized handling required, so no meaningful cost comparison favors AI; a trained technologist is still required. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can perform the physical, in-lab aspects of radioactive substance preparation and application to biological specimens. This remains a task requiring human technician presence and expertise in a controlled laboratory setting. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs radioimmunoassay-style physical specimen preparation; this remains a manual laboratory task requiring human dexterity and safety protocols. |
Administer radiopharmaceuticals or radiation intravenously to detect or treat diseases, using radioisotope equipment, under direction of a physician.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.9/5 · click for rater detail
Administer radiopharmaceuticals or radiation intravenously to detect or treat diseases, using radioisotope equipment, under direction of a physician.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No adoption of AI for this task is occurring because it is fundamentally a regulated, hands-on clinical procedure. The regulatory framework and safety requirements mean this task will remain human-performed. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare imaging/radiopharmacy is a highly regulated, physically-intensive clinical setting with minimal AI adoption for hands-on procedures like this. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist marginally with dose calculations, patient record preparation, or equipment calibration decision support, but the core task of intravenous administration is human-dependent. Limited augmentation value for the primary procedural work. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI may assist with dosage calculations, scheduling, or documentation, but offers little direct support for the physical act of intravenous administration itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires direct physical administration of radioactive materials intravenously—a procedural, hands-on action that current AI systems cannot perform. No end-to-end automation is feasible for the injection itself, and the safety-critical nature precludes meaningful time savings via AI alone. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on invasive medical procedure requiring physical dexterity, patient assessment, and IV administration that current AI systems cannot physically perform. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and regulatory barriers exist: only licensed nuclear medicine technologists can administer radiopharmaceuticals under state and federal regulations (NRC, state medical boards). Physician oversight is mandated, and liability for radiation administration is legally assigned to qualified humans. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Administering radioactive substances intravenously requires licensed, certified personnel under physician direction, with strict regulatory and safety requirements making human performance legally mandatory. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI cannot perform this task at all, making any cost comparison moot. The loaded cost of a nuclear medicine technologist far exceeds any theoretical AI cost for a task AI cannot execute. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical task, so cost comparison is moot and the human remains the only viable option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can physically administer intravenous radiopharmaceuticals. This remains entirely a human clinical task performed by licensed technologists under physician supervision. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product administers radiopharmaceuticals intravenously; this remains entirely a manual clinical task performed by trained technologists. |
Prepare stock radiopharmaceuticals, adhering to safety standards that minimize radiation exposure to workers and patients.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail
Prepare stock radiopharmaceuticals, adhering to safety standards that minimize radiation exposure to workers and patients.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare automation in high-stakes nuclear medicine is extremely conservative; no measurable displacement of radiopharmaceutical preparation by AI exists, and regulatory structure actively prevents it. Adoption remains at zero in production settings. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare radiopharmacy/nuclear medicine handling is a highly regulated, physically-grounded niche with minimal AI-driven automation of the physical preparation process. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could modestly assist with record-keeping, inventory tracking, or documentation of safety protocols, but the core task of handling radioactive materials and making real-time safety decisions offers limited scope for meaningful augmentation while the human remains in the loop. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with dose calculation, inventory tracking, or documentation support, but offers little help with the core physical mixing/preparation and radiation safety compliance steps. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Preparing radiopharmaceuticals requires precise physical manipulation of radioactive materials in a controlled environment, real-time safety assessment, and compliance with strict nuclear regulatory protocols. Current AI systems cannot physically handle hazardous materials, adjust dosages in real-time based on equipment feedback, or take responsibility for radiation safety decisions. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical task involving handling radioactive materials, calibrating doses, and following strict safety protocols; no AI system can physically prepare radiopharmaceuticals today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task is protected by multiple hard barriers: nuclear medicine technologists must be licensed by state boards, the task is directly regulated by the NRC and state radiation control programs, and federal law requires a licensed human to prepare radiopharmaceuticals. Liability for radiation errors is substantial and legally non-delegable. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Radiopharmaceutical handling requires licensed technologists under strict regulatory and radiation safety oversight (NRC/state licensing), with legal requirements for certified personnel to perform this task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task involves handling expensive radiopharmaceutical materials and requires licensed personnel with years of training; any AI system capable of the physical and regulatory requirements would cost far more than the technologist's fully-loaded compensation. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute for the physical handling and preparation work, so AI cost is not comparable—human labor is the only option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product or system can independently prepare stock radiopharmaceuticals; this task remains strictly within human-supervised, licensed technologist domain. The regulatory framework (NRC, state boards) mandates human licensure and direct accountability, not AI operation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical radiopharmaceutical preparation; this remains a manual clinical task performed by trained, licensed technologists. |
Dispose of radioactive materials and store radiopharmaceuticals, following radiation safety procedures.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail
Dispose of radioactive materials and store radiopharmaceuticals, following radiation safety procedures.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of automation in radioactive material handling is minimal; regulatory requirements, safety criticality, and the need for human accountability create structural barriers that prevent even exploration of AI-driven solutions in practice. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare physical/clinical tasks involving hazardous materials handling show minimal AI adoption; this is a low-digitization, hands-on task. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While AI-assisted monitoring systems (alerts, sensor integration, documentation) could provide marginal assistance, the core task—physical handling and real-time safety decisions—remains fundamentally human, and augmentation offers limited productivity gains. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with logging, inventory tracking, or compliance documentation, but offers little help with the physical disposal and storage actions themselves. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires precise physical manipulation of hazardous radioactive materials in controlled environments, real-time radiation monitoring, and compliance with strict safety protocols that demand human judgment and immediate adaptation to unexpected conditions. Current AI systems lack the embodied capability, safety certification, and legal authority to handle radioactive substances. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physical handling of hazardous radioactive materials, secure storage, and compliance documentation—AI systems today cannot physically manipulate or store these materials. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task is subject to hard legal and regulatory barriers: only licensed Nuclear Medicine Technologists or equivalent radiation workers can legally dispose of or store radioactive materials under NRC and state regulations, and liability for improper handling is severe. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Radiation safety and materials handling are governed by strict NRC/state licensing, certification, and legal chain-of-custody requirements mandating qualified human technologists. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of developing, certifying, and maintaining specialized robotic systems capable of handling radioactive materials, plus liability insurance and regulatory compliance, vastly exceeds the loaded wage of a nuclear medicine technologist performing the task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task, so AI cost is not comparable—human labor is the only option currently. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products perform autonomous disposal or storage of radioactive materials in production; regulatory frameworks (NRC, state radiation control boards) require licensed human technologists to perform these tasks, and no commercial system has demonstrated reliable autonomous execution of this safety-critical function. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs physical radioactive waste disposal or storage; this remains entirely manual, regulated human work. |
Measure glandular activity, blood volume, red cell survival, or radioactivity of patient, using scanners, Geiger counters, scintillometers, or other laboratory equipment.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.4/5 · click for rater detail
Measure glandular activity, blood volume, red cell survival, or radioactivity of patient, using scanners, Geiger counters, scintillometers, or other laboratory equipment.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Nuclear medicine is a highly regulated, small specialty field with slow digitization. Institutions invest in certified human technologists because regulatory and safety requirements mandate human expertise and accountability; there is minimal incentive or pathway for automation in this domain. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Nuclear medicine technology is a highly physical, equipment-dependent clinical field with minimal AI agent deployment for the hands-on measurement components. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI might assist in image interpretation or post-measurement analysis of scan results, but offers minimal augmentation to the core measurement task itself—positioning equipment, taking readings, and ensuring proper patient positioning and safety requires human judgment and physical presence. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with data logging, dose calculations, or interpreting scan outputs, but offers little direct help with the physical measurement process itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires direct physical measurement of patient radioactivity using specialized laboratory equipment (scanners, Geiger counters, scintillometers) that must be positioned and operated in situ. No current AI system can autonomously operate these physical instruments or position them on patients without human intervention. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physically operating radiation detection equipment, positioning patients, and handling radioactive materials—none of which current AI systems can perform end-to-end without a human physically present. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Nuclear medicine technologists must be licensed/certified in most jurisdictions and are required to handle radioactive materials under strict regulatory oversight (NRC, state boards). A human licensed technologist must legally perform or sign off on patient measurement and radiation safety protocols. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Handling radioactive materials and operating medical imaging equipment on patients requires licensure, radiation safety certification, and direct human accountability, making substitution legally and physically barred. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The equipment, facility, specialized training, and regulatory compliance costs are high regardless of automation. The human technologist's wage is modest relative to the capital and overhead, making any attempted automation of the measurement operation more expensive than retaining the skilled technician. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical measurement task, so cost comparison favors the human technologist by default. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | While AI can analyze resulting scan images offline, no deployed product performs the hands-on measurement and equipment operation that the task statement explicitly requires. The actual measurement step—handling equipment and taking readings on living patients—remains manual. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously measures glandular activity or radioactivity using scanners or Geiger counters on live patients; this remains fundamentally a hands-on clinical task. |
Train or supervise student or subordinate nuclear medicine technologists.
0CI 0–0 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail
Train or supervise student or subordinate nuclear medicine technologists.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare remains among the slowest sectors for AI agent adoption in safety-critical roles; training and supervision of technologists in a nuclear environment is unlikely to see significant AI displacement given regulatory, liability, and professional credentialing barriers. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare training and supervision functions show minimal AI adoption for direct replacement, given regulatory and safety constraints in clinical education. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could potentially assist a technologist trainer by organizing curricula, flagging gaps in competency assessments, or summarizing procedure guidelines, but the core mentoring, observation, and judgment remain solidly human-dependent with limited augmentation upside. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can support training via simulations, quizzes, or supplementary materials, but the core supervisory and mentorship task still needs a human trainer. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Training and supervision of technologists requires ongoing real-time interaction, adaptive feedback, assessment of competency, and nuanced judgment about individual learning needs and clinical readiness. Current AI systems cannot replicate the dynamic mentoring relationship or provide credible hands-on oversight in a regulated medical environment. |
| Task automatability | claude-sonnet-5 | 1/5 | Training and supervising trainees requires hands-on clinical demonstration, real-time feedback, and human mentorship that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Nuclear medicine is a highly regulated field (NRC, state licensing, accreditation standards) where human oversight and sign-off by qualified personnel is typically mandated. Legal and liability requirements strongly require a licensed human technologist to train and certify subordinates. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Supervision and training of licensed medical technologists typically requires certified, experienced personnel per regulatory and accreditation standards, making this a hard professional barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | An experienced nuclear medicine technologist conducting training and supervision commands a professional salary; the liability and compliance overhead of deploying AI for this task would far exceed any cost savings, especially given regulatory and safety criticality. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for this supervisory role, so cost comparison favors humans by default. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs live training, real-time supervision, or competency assessment of medical technologists in production settings. While AI can deliver static content or assist with documentation, it cannot supervise someone performing safety-critical nuclear medicine procedures. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product supervises or trains nuclear medicine technologists in clinical settings; this remains entirely a human responsibility. |
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.