The first study to map radiologists and scanners together
The US radiologist shortage is well documented. What's been missing is a clean picture of where that shortage actually leaves patients without any local imaging option at all — because geographic analyses of the radiologist workforce and of imaging equipment have historically been done separately. A national county-level study published October 2, 2026 in Academic Radiology, led by Adyasha Pradhan, a medical student at the Touro College of Osteopathic Medicine in New York, with co-authors Sharon D'Souza and Andrew B. Rosenkrantz, closes that gap by combining Medicare and National Plan and Provider Enumeration System (NPPES) data into a single county-level typology.
The reason the distinction matters: imaging acquisition and interpretation have been decoupled by teleradiology. "Teleradiology allows a county with imaging equipment and no identified locally practicing radiologist to receive remote interpretation of the studies its equipment can acquire," Pradhan and co-authors wrote, as reported by Radiology Business. "A county with neither equipment nor radiologists receives no local imaging at all, and its residents must travel for even basic studies."
Four Americas, by imaging access
About 54% of US counties had no documented local radiologist as of June 2026. That statistic alone makes the shortage sound almost universal — but it collapses two very different situations into one number. The study splits counties into four categories:
| County type | Definition | Share of counties | Population affected |
|---|---|---|---|
| Fully served | Local radiologist and local imaging equipment | 46% | 307.4M residents (~90%) |
| Equipment-only desert | Scanner present, no local radiologist — relies on teleradiology | 36% | 25.9M residents (~8%) |
| True imaging desert | No local scanner and no local radiologist | 18% | 6.4M residents (~2%) |
| Radiologist-only desert | Local radiologist, no local imaging equipment | 19 counties | ~400,000 residents (~0.1%) |
The equipment-only deserts are the ones teleradiology already serves, however imperfectly. The 568 counties the study classifies as true imaging deserts are the harder problem: there's no scanner to connect a remote reader to in the first place, so residents have to travel for even a basic CT or X-ray.
The gap shows up in how much care people actually get
This isn't just a geography curiosity — it shows up in utilization. Medicare beneficiaries living in true imaging deserts had about 6% fewer imaging events and 9% lower imaging spending than beneficiaries in fully served counties, per the study. Fewer studies and less spending in a population that is, on average, no healthier than its neighbors is a sign of unmet need, not efficient care.
The subspecialty picture is starker still. All five major subspecialties the researchers tracked — interventional, neuroradiology, nuclear medicine, pediatric radiology and body imaging — coexisted locally in only about 6% (194) of US counties. About 80% of counties had no locally designated vascular/interventional radiologist, representing 81.5 million residents (24% of the population); 86% had no local neuroradiologist, 87% no local body imager, 89% no local nuclear medicine specialist, and 92% no local pediatric radiologist. General radiologists cover a meaningful share of that subspecialty-equivalent reading, the authors noted, but for image-guided interventional procedures specifically, there's no remote substitute — a patient still has to be where the equipment and the proceduralist both are.
Two different problems need two different fixes
The authors argue equipment-only deserts and true imaging deserts call for distinct remedies. For the 36% of counties that already have a scanner but no local radiologist, the gap is a staffing and coverage problem that remote interpretation already addresses in part; the study points to loan repayment programs, rural residency expansion, and expanded visa waiver pathways as ways to grow the supply of radiologists willing to practice in or cover those areas.
True imaging deserts are a capital problem first. "If capital investment in local scanners is unavailable, then greater development of organized transport and referral pathways to neighboring counties may be warranted," the authors wrote. In other words: where buying a scanner isn't realistic, the fallback is making the trip to one easier, not eliminating the trip.
What it takes to stand up CT capability without a local radiologist
The study doesn't evaluate AI-assisted reporting — its policy menu is deliberately about workforce and capital. But its own framing points at the practical bottleneck for any rural site weighing whether to add a scanner: acquisition and interpretation are decoupled, so a county doesn't need a radiologist to live there, it needs one, somewhere, willing and able to read its studies affordably and on a timeline that works. For a hospital or imaging center converting from an equipment-only desert into a genuinely served site, that reading arrangement is the whole decision.
Traditional nighthawk and teleradiology contracts solve the coverage problem but not always the economics of a low-volume site: many carry minimum-volume commitments, per-study rates that assume higher throughput, and onboarding timelines measured in months — a hard case to make at a facility that might read a few dozen CTs a week, not a few hundred. That's the gap AI-assisted CT reporting is built to narrow. By producing a complete, structured report draft for a radiologist to review rather than dictate from scratch, it lets a given remote reading radiologist cover more studies per hour, which lowers the effective per-study cost of contracting that coverage and shortens how long it takes to bring a new reading relationship online — changing the math on whether a scanner pencils out in a county that could never justify the traditional staffing model.
That's where xAID fits into this specific gap: a foundation model drafts the structured report, xAID's in-house radiologist reviews every preliminary, and the facility's contracted or remote reading radiologist gets a ready-to-sign report rather than a blank worklist — the same decoupled acquisition-and-interpretation model the study describes, built to make covering a lower-volume site economically workable.
Frequently asked questions
What is a radiology imaging desert?
A radiology "imaging desert" is a county with neither a locally practicing radiologist nor local imaging equipment, so residents must travel elsewhere for even basic studies. A national county-level study published October 2, 2026 in Academic Radiology found about 18% of US counties meet this definition, touching roughly 2% of the population, or 6.4 million people.
How many US counties have no local radiologist at all?
About 54% of US counties had no documented local radiologist as of June 2026, according to the study's analysis of Medicare and National Plan and Provider Enumeration System data. Most of those counties — 36% of all US counties, representing about 25.9 million residents — still have local imaging equipment and rely on teleradiology for interpretation; the rest are true imaging deserts with no equipment either.
Does living in an imaging desert change how much imaging care people get?
Yes. The study found Medicare beneficiaries living in true imaging deserts had about 6% fewer imaging events and 9% lower imaging spending than beneficiaries in "fully served" counties with both local radiologists and equipment — evidence that the access gap translates into less care, not just less convenient care.
Can a rural facility add CT capability without an on-site or regional radiologist?
Yes, if it can line up remote interpretation. A scanner plus a teleradiology or remote-reading contract is what already defines the 36% of US counties the study classifies as equipment-only deserts. The practical bottleneck is usually economics and timeline: a low-volume site needs a remote-reading arrangement that is affordable at its volume and fast to stand up, which is where AI-assisted report drafting — producing a structured, ready-to-sign draft for a remote radiologist to review — can change the math versus waiting on a traditional nighthawk contract alone.
Source: Pradhan A, D'Souza S, Rosenkrantz AB. "A National County-Level Typology of Imaging Infrastructure and the Radiologist Workforce in the United States." Academic Radiology, October 2, 2026. doi.org/10.1016/j.acra.2026.09.030; as reported by Radiology Business. Figures are rounded as reported.