Nuclear Medicine Physicians

Prepare comprehensive interpretive reports of findings.

AI: Fully automatable - By 2025 AI systems can analyze nuclear medicine images and generate comprehensive interpretive reports of findings at a clinically useful level, though clinician review is commonly retained.

Compare nuclear medicine procedures with other types of procedures such as computed tomography, ultrasonography, nuclear magnetic resonance imaging, and angiography.

AI: Fully automatable - Comparing nuclear medicine procedures with CT, US, MRI, and angiography is primarily an information‑synthesis task that AI can perform fully using physics, clinical indications, and literature synthesis.

Provide advice on the selection of nuclear medicine supplies or equipment.

AI: Fully automatable - AI can fully analyze specifications, costs, compatibility, and evidence to provide well-founded recommendations for selecting nuclear medicine supplies and equipment.

Schedule examinations and staff activities.

AI: Fully automatable - Scheduling examinations and staff activities are routine administrative tasks that AI systems can automate end-to-end, including optimization and conflict resolution.

Check and approve the quality of diagnostic images before patients are discharged.

AI: Partial - AI can reliably assess image quality and flag issues automatically, but final approval before patient discharge typically remains a human responsibility and may require contextual judgment.

Establish and enforce radiation protection standards for patients and staff.

AI: Partial - AI can propose radiation protection protocols, optimize dosing and monitoring, and detect breaches, but cannot assume institutional authority to set policy or enforce regulations without human oversight.

Interpret imaging data and confer with other medical specialists to formulate diagnoses.

AI: Partial - AI can interpret imaging and produce differential diagnoses and reports, but multidisciplinary discussion, clinical integration, and final diagnostic responsibility require human specialists.

Prescribe radionuclides and dosages to be administered to individual patients.

AI: Partial - AI can calculate radionuclide selection and patient‑specific dosages based on protocols and patient data, but legal prescribing and final dose authorization remain clinician responsibilities.

Review procedure requests and patients' medical histories to determine applicability of procedures and radioisotopes to be used.

AI: Partial - AI can analyze procedure requests and EHR data and recommend applicable procedures and radioisotopes, but cannot assume final clinical responsibility or fully account for nuanced, context-specific judgement and regulatory constraints.

Determine appropriate tests or protocols based on patients' needs or conditions.

AI: Partial - AI can propose appropriate tests or protocols based on guidelines and patient data, yet complex clinical judgment and individualized decision‑making still require physician oversight.

Direct nuclear medicine technologists or technicians regarding desired dosages, techniques, positions, and projections.

AI: Partial - AI can generate dosing, technique, positioning and projection recommendations for technologists, but cannot legally or practically replace a human physician’s direct supervision and hands‑on adjustments.

Monitor quality control of radionuclide preparation, administration, or disposition ensuring that activities comply with applicable regulations and standards.

AI: Partial - AI can automate many QC measurements and flag regulatory nonconformities in radionuclide preparation and administration, but human verification, legal signoff, and resolution of complex QC issues remain necessary.

Monitor handling of radioactive materials to ensure that established procedures are followed.

AI: Partial - AI systems can monitor sensor feeds and flag deviations in handling radioactive materials, but final compliance enforcement and complex situational responses require human oversight and accountability.

Advise other physicians of the clinical indications, limitations, assessments, or risks of diagnostic and therapeutic applications of radioactive materials.

AI: Partial - AI can produce evidence‑based advisories about indications, limitations, and risks, but peer communication, contextual nuance, and medico‑legal responsibility limit fully autonomous advisory roles.

Interview and physically examine patients prior to testing.

AI: Partial - AI can conduct comprehensive history‑taking and symptom screening remotely, but cannot perform the physical examination components that require direct tactile assessment and clinical interpretation.

Calculate, measure, or prepare radioisotope dosages.

AI: Partial - AI can calculate dose values and provide preparation protocols, but cannot physically measure or legally prepare radioactive materials or take responsibility for handling.

Teach nuclear medicine, diagnostic radiology, or other specialties at graduate educational level.

AI: Partial - AI can generate lectures, assessments, and teaching materials and support learning at the graduate level, but cannot fully replace human faculty for clinical supervision, mentorship, and accreditation responsibilities.

Consult with patients following radiation treatments to provide information and assess outcomes or to recommend further consultation or treatments as appropriate.

AI: Partial - AI can provide information, symptom triage, and preliminary outcome assessment, but cannot perform full clinical examinations, provide legally accountable medical advice, or replace in-person patient consultations.

Monitor cleanup of radioactive spills to ensure that proper procedures are followed and that decontamination activities are conducted.

AI: Partial - AI can recommend spill response steps, model contamination spread, and monitor sensor data remotely, but cannot fully supervise or execute on-site decontamination and regulatory incident command roles.

Test dosage evaluation instruments and survey meters to ensure they are operating properly.

AI: Partial - AI can supply calibration procedures, interpret test data, and guide technicians, but cannot physically test or certify instrument operation without human or robotic execution and oversight.

Formulate plans and procedures for nuclear medicine departments.

AI: Partial - AI can draft comprehensive departmental plans and standard operating procedures using best practices and regulations, but final tailoring, approval, and accountability require human leadership.

Direct the safe management and disposal of radioactive substances.

AI: Partial - AI can provide guidance on regulatory-compliant disposal methods and optimize management workflows, but cannot assume the legal and operational authority to direct physical handling and disposal activities.

Consult with anesthesiologists regarding recommended dosages or combinations of sedative drugs.

AI: Partial - AI can compute guideline-based sedative dosing options and interactions to inform anesthesiologists, but cannot replace clinician judgment or assume responsibility for individualized dosing decisions.

Conduct laboratory procedures, such as radioimmunoassay studies of blood or urine, using radionuclides.

AI: Partial - Laboratory radionuclide assays are widely automated and AI can control/interpret them, but handling, licensing, and QA currently require human oversight and intervention.

Perform cardiovascular nuclear medicine procedures such as exercise testing and pharmacologic stress testing.

AI: Not automatable - Performing exercise and pharmacologic stress testing requires physical patient interaction, real‑time physiological monitoring and immediate clinical judgment that AI cannot autonomously provide as of 2025.

Administer radioisotopes to clinical patients or research subjects.

AI: Not automatable - Administering radioisotopes involves hands‑on procedures, sterile technique and regulated licensing that AI cannot perform or be legally authorized to carry out in 2025.