Role Definition
| Field | Value |
|---|---|
| Job Title | Interventional Radiologist |
| Seniority Level | Mid-to-Senior (board-certified, 5+ years post-fellowship) |
| Primary Function | Physician who performs minimally invasive, image-guided procedures -- threading catheters through blood vessels under fluoroscopy, CT, or ultrasound to treat vascular disease (angioplasty, stenting, embolisation), perform tumour ablation, place drainage catheters, conduct biopsies, insert ports and lines, and manage acute haemorrhage. Works in angiography suites, hybrid operating rooms, and interventional oncology units. Treats patients directly, manages complications intra-procedurally, and provides longitudinal clinical follow-up. |
| What This Role Is NOT | Not a Diagnostic Radiologist (reads images from a workstation -- no procedures, no patient contact during treatment). Not an Interventional Cardiologist (focuses on coronary arteries and cardiac catheterisation). Not a Vascular Surgeon (performs open surgical procedures). Not a Radiation Oncologist (delivers external beam radiation). The distinction matters: diagnostic radiology is primarily cognitive image interpretation; interventional radiology is hands-on procedural medicine. |
| Typical Experience | MD/DO + 5-year diagnostic radiology residency + 1-2 year IR fellowship (or 6-year integrated IR residency). ABR dual board certification in diagnostic and interventional radiology. State medical licence + DEA registration. Typically 5-20+ years post-training. Many hold subspecialty expertise in interventional oncology, neurointerventional, or paediatric IR. |
Seniority note: Seniority does not materially change the zone. All board-certified IRs perform the same core procedural work -- catheter navigation, embolisation, ablation, drainage. Senior IRs take more complex cases (portal hypertension, paediatric, neuro) and departmental leadership roles, which are equally or more AI-resistant.
- Protective Principles + AI Growth Correlation
| Principle | Score (0-3) | Rationale |
|---|---|---|
| Embodied Physicality | 3 | IRs physically thread guidewires and catheters through vasculature, manipulate microcatheters to sub-millimetre precision under live fluoroscopy, place stents, perform radiofrequency and microwave ablation, insert chest drains, and manage procedural complications with hands on instruments in a sterile field. Every case is a manual procedure in unpredictable patient anatomy. |
| Deep Interpersonal Connection | 2 | IRs obtain informed consent, manage peri-procedural patient anxiety (often with conscious sedation, not general anaesthesia -- the patient is awake), coordinate with referring clinicians and anaesthesia teams, and provide longitudinal follow-up for oncology and vascular patients. Less longitudinal than primary care but intense and trust-dependent. |
| Goal-Setting & Moral Judgment | 3 | IRs independently decide procedural approach (which vessel to access, which embolic agent, whether to abort a procedure when risk exceeds benefit), titrate conscious sedation, manage intra-procedural complications (vessel perforation, non-target embolisation), and bear full medical-legal accountability for procedural outcomes. Autonomous physician-level clinical judgment in real time. |
| Protective Total | 8/9 | |
| AI Growth Correlation | 0 | AI adoption neither creates nor destroys IR demand. Demand is driven by aging population, rising cancer incidence (IR ablation/embolisation), shift from open surgery to minimally invasive approaches, and acute workforce shortage -- not AI deployment. |
Task Decomposition (Agentic AI Scoring)
| Task | Time % | Score (1-5) | Weighted | Aug/Disp | Rationale |
|---|---|---|---|---|---|
| Catheter-based vascular interventions -- angioplasty, stenting, embolisation, thrombolysis | 30% | 1 | 0.30 | NOT INVOLVED | Threading catheters through tortuous vasculature, navigating guidewires past stenoses, deploying stents, and delivering embolic agents requires continuous manual dexterity in unique patient anatomy under live fluoroscopy. CorPath GRX (Siemens) provides robotic assistance for coronary PCI but remains operator-controlled, has limited IR vascular adoption, and cannot handle the anatomical variability of peripheral, visceral, and neuro-interventional cases. No autonomous catheter navigation exists. |
| Tumour ablation and interventional oncology -- RFA, MWA, cryoablation, TACE, Y-90 | 15% | 2 | 0.30 | AUGMENTATION | AI tumour segmentation tools assist pre-procedural planning (automated volume calculation, ablation margin prediction). Intra-procedurally, the IR physically places ablation probes under imaging guidance, manages real-time thermal monitoring, and adapts to tumour movement and patient physiology. AI augments planning; the proceduralist executes. |
| Image-guided biopsies and drainage -- percutaneous biopsy, abscess drainage, chest drain insertion | 10% | 2 | 0.20 | AUGMENTATION | AI-assisted needle tracking and image fusion tools improve targeting accuracy. The IR physically advances the needle, adjusts trajectory in real time based on patient breathing and anatomy, and manages complications (pneumothorax, bleeding). No autonomous needle placement system is FDA-approved or in clinical use. |
| Pre-procedural planning and case review -- imaging interpretation, treatment planning, MDT discussion | 10% | 3 | 0.30 | AUGMENTATION | AI tools for diagnostic image interpretation, 3D vascular reconstruction, and automated tumour measurement assist planning. IRs interpret imaging with domain-specific procedural intent (access route, device selection, complication risk) that generic diagnostic AI does not replicate. The 3+ score reflects diagnostic AI tool maturity applied to the planning component. |
| Intra-procedural monitoring and imaging -- fluoroscopy, cone-beam CT, ultrasound guidance | 15% | 2 | 0.30 | AUGMENTATION | AI-enhanced image quality tools (noise reduction, dose optimisation, automated cone-beam CT reconstruction) improve real-time imaging. The IR interprets images in procedural context, adjusts technique, and makes real-time decisions based on what the imaging reveals mid-procedure. AI improves the image; the human acts on it. |
| Complication management and emergency response -- vessel perforation, non-target embolisation, haemorrhage | 10% | 1 | 0.10 | NOT INVOLVED | When a vessel perforates or embolic material migrates, the IR must immediately redirect catheters, deploy rescue devices, or convert to emergency measures. Unpredictable, high-stakes, requires manual dexterity and crisis judgment. No AI involvement. |
| Documentation, billing, and administrative | 10% | 4 | 0.40 | DISPLACEMENT | Procedural documentation increasingly automated by AI-populated structured reports, voice-to-text, and AIMS integration. IR reviews and signs. |
| Total | 100% | 1.90 |
Task Resistance Score: 6.00 - 1.90 = 4.10/5.0
Displacement/Augmentation split: 10% displacement, 50% augmentation, 40% not involved.
Evidence Score
| Dimension | Score (-2 to 2) | Evidence |
|---|---|---|
| Job Posting Trends | 2 | BLS projects 3% physician growth 2024-2034. IR-specific growth estimated at 7% 2023-2033 (Research.com 2026), faster than physician average, driven by minimally invasive procedure adoption. IR is among top 5 specialties with critical workforce shortages (SIR/ACR Joint Task Force). RSNA 2025 survey: more than half the world lacks IR access. AAMC projects radiologist shortage of 37,800-124,000 by 2034. |
| Company Actions | 2 | No health system is cutting IR positions citing AI. Hospitals are expanding IR suites and recruiting aggressively. Small and rural practices face acute IR recruitment challenges (SIR/ACR 2025). Forbes (Jan 2026): radiologist demand grew 17% despite AI permeation in imaging. CNN (Feb 2026): radiology has become the "case study for why AI won't replace human workers." |
| Wage Trends | 2 | Average IR salary $400,000-$550,000 (Research.com 2026, Salary.com). Doximity 2025 reports ~$571,700. Private practice IRs reaching $600,000-$800,000 (AllStar Healthcare 2025). IR earns 15-30% premium over diagnostic radiology. Salaries outpacing inflation driven by shortage economics. |
| AI Tool Maturity | 1 | 1,247 FDA-authorised AI medical devices as of Aug 2025, 75%+ in radiology -- but overwhelmingly diagnostic image interpretation. For IR specifically: AI tumour segmentation and ablation planning tools in clinical use. CorPath GRX provides robotic-assisted catheter manipulation but remains fully operator-controlled. No autonomous procedural AI in IR. SHERPA consortium (EU, Mar 2026) just initiated clinical studies for AI-assisted minimally invasive brain/cancer treatments -- early stage. |
| Expert Consensus | 2 | Broad consensus that IR is highly AI-resistant due to procedural nature. Lastrucci (2025, MDPI): "full autonomy in IR procedures remains distant." Huth (2025, JVIR): AI will reshape IR workflow but the proceduralist remains central. Friebe (2025): AI improves probe localisation and ablation planning, does not replace the operator. Forbes (Jan 2026): radiologist demand grew 17% despite AI -- IR even more protected than diagnostic radiology given procedural nature. CNN (Feb 2026): radiology is the "case study" for why AI won't replace workers. No credible expert predicts IR displacement. |
| Total | 9 |
Barrier Assessment
| Barrier | Score (0-2) | Rationale |
|---|---|---|
| Regulatory/Licensing | 2 | MD/DO + 5-6 year radiology residency + 1-2 year IR fellowship + ABR dual board certification + state medical licence + DEA registration. No regulatory pathway for AI as independent proceduralist. FDA has not approved any autonomous interventional system for IR procedures. |
| Physical Presence | 2 | IRs must be physically present in the angiography suite for every procedure. Threading catheters, placing stents, performing ablation, and managing complications are irreducible manual tasks. No telemedicine pathway for interventional procedures. Robotic-assisted systems (CorPath GRX) still require the physician at the control console in the same room. |
| Union/Collective Bargaining | 1 | Physicians are not significantly unionised. However, IR is a small specialty (~5,000-6,000 active IRs in the US per SIR) with strong professional society advocacy (SIR, ACR). The small workforce creates natural supply-side protection against displacement. |
| Liability/Accountability | 2 | IRs carry personal malpractice liability for every procedure. Complications (vessel perforation, non-target embolisation, organ injury) create substantial legal exposure. No insurer or hospital would accept "the AI navigated the catheter" as a liability defence. DEA accountability for controlled substances administered during conscious sedation. |
| Cultural/Ethical | 2 | Patients and referring physicians fundamentally expect a human physician to physically perform invasive procedures inside their body. The concept of an autonomous robot threading a catheter through a patient's femoral artery into their liver without a physician's hands is culturally and ethically unacceptable for the foreseeable future. |
| Total | 9/10 |
AI Growth Correlation Check
Confirmed 0 (Neutral). AI adoption does not create or destroy IR demand. Demand drivers are structural: aging population (increasing vascular disease and cancer), global shift from open surgery to minimally invasive approaches, workforce shortage (top 5 physician shortage specialty, >50% of world lacks IR access), and expanding clinical indications (IR is absorbing procedures formerly performed by surgeons). Not Accelerated Green -- no recursive AI dependency.
JobZone Composite Score (AIJRI)
| Input | Value |
|---|---|
| Task Resistance Score | 4.10/5.0 |
| Evidence Modifier | 1.0 + (9 x 0.04) = 1.36 |
| Barrier Modifier | 1.0 + (9 x 0.02) = 1.18 |
| Growth Modifier | 1.0 + (0 x 0.05) = 1.00 |
Raw: 4.10 x 1.36 x 1.18 x 1.00 = 6.5780
JobZone Score: (6.5780 - 0.54) / 7.93 x 100 = 76.2/100
Zone: GREEN (Green >=48, Yellow 25-47, Red <25)
Sub-Label Determination
| Metric | Value |
|---|---|
| % of task time scoring 3+ | 10% |
| AI Growth Correlation | 0 |
| Sub-label | Green (Stable) -- <20% task time scores 3+, Growth Correlation not >=1 |
Assessor override: None -- formula score accepted. Score of 76.2 places IR above Anesthesiologist (73.8) in the Green Stable tier. The 2.4-point gap reflects IR's stronger barrier protection (9/10 vs 8/10 -- no telemedicine pathway, smaller workforce creating natural supply-side protection) and higher proportion of score-1 tasks (40% vs 25%). Both roles share maximum physicality and moral judgment scores (3/3 each). Higher than Anesthesiologist is correct: 40% of IR task time has zero AI involvement, compared to 25% for anesthesiology.
Assessor Commentary
Score vs Reality Check
The 76.2 score and Green (Stable) label are honest. IR sits 28.2 points above the Green/Yellow boundary at 48. Only 10% of task time (documentation) faces displacement. 50% is augmented and 40% is entirely untouched by AI. The "Stable" sub-label is correct: daily procedural work -- catheter navigation, embolisation, ablation, drainage, complication management -- has no AI substitute and will not change materially over the next decade. The critical distinction from diagnostic radiology (which faces more AI augmentation pressure on its core image-reading function) is that IR's value is in the hands, not the eyes.
What the Numbers Don't Capture
- Robotic-assisted IR is real but early. CorPath GRX (Siemens Healthineers) is FDA-cleared for robotic-assisted PCI and has been trialled in neurovascular cases (Pereira 2023, 50 citations). It reduces radiation exposure to the operator by 20% and improves precision -- but it is fully physician-controlled, not autonomous. The trajectory is toward surgeon-controlled robotic assistance (analogous to da Vinci in surgery), not physician replacement.
- Diagnostic radiology AI spillover. As AI handles more routine diagnostic image interpretation, radiologists may increasingly shift toward procedural work (IR) to maintain clinical relevance. This could increase competition for IR training slots but reinforces the specialty's protected status.
- The diagnostic-interventional gap is widening. AI augments diagnostic radiology's core function (image reading) at scale. It barely touches IR's core function (physical procedures). This gap will widen as diagnostic AI matures, making IR increasingly the "irreducible core" of radiology as a medical discipline.
- Global shortage is structural. The RSNA 2025 survey finding that >50% of the world population lacks IR access is not a temporary market fluctuation -- it reflects the 13-15 year training pipeline, limited fellowship slots, and the specialty's relative obscurity outside academic centres.
Who Should Worry (and Who Shouldn't)
IRs performing complex catheter-based procedures in angiography suites -- embolisation, TIPS, complex stenting, tumour ablation, emergency haemorrhage control -- are the safest version of this role. Every case combines manual catheter manipulation, real-time imaging interpretation in procedural context, and crisis-ready judgment. Subspecialists in interventional oncology and neurointerventional radiology are particularly protected -- these involve the highest complexity and most unpredictable anatomy.
IRs whose practice has shifted primarily toward simple, repetitive image-guided procedures (routine biopsies, line placements) should pay moderate attention -- these structured tasks are the most likely candidates for eventual robotic assistance. IRs spending significant time on diagnostic image interpretation rather than procedures should pay more attention -- that cognitive work faces the same AI augmentation pressure as pure diagnostic radiology.
The single biggest separator: whether you are physically standing at the angiography table navigating catheters inside a patient. If you are, you are among the most AI-resistant physicians in medicine.
What This Means
The role in 2028: IRs will use AI-enhanced pre-procedural planning (automated tumour segmentation, 3D vascular mapping, ablation margin prediction) as standard workflow. Robotic-assisted catheter systems will expand from cardiology into peripheral vascular and neuro-interventional cases, functioning as precision-enhancing tools under physician control. Core procedural work -- catheter navigation, embolisation, ablation, complication management -- remains entirely human. IR workforce shortage continues to worsen globally.
Survival strategy:
- Pursue subspecialty depth in interventional oncology or neurointerventional radiology -- the most complex, least automatable, and highest-compensated IR work
- Embrace robotic-assisted platforms (CorPath GRX, emerging systems) as precision tools that reduce radiation exposure and improve outcomes -- position yourself as the physician who masters human-robot collaboration in the IR suite
- Maintain and develop the procedural skill set that defines the role -- catheter manipulation, complication management, emergency haemorrhage control -- because these manual skills are the moat that separates IR from every AI capability on the horizon
Timeline: 20+ years. Driven by the convergence of irreducible manual procedures (catheter navigation, ablation probe placement), no FDA pathway for autonomous interventional systems, personal medical-legal liability for invasive procedures, fundamental cultural expectation of a human physician performing surgery inside your body, and a worsening global workforce shortage with a 13-15 year training pipeline.
