Radiology's capacity problem, and the levers that actually move it
Mike Moore
·
September 8, 2026
A companion piece to the Flight Path conversation with Jessica Bordeaux, CEO of Raleigh Radiology.
The radiologist shortage is usually described as a hiring problem. Jessica Bordeaux, the CEO of Raleigh Radiology, would tell you that framing is the trap. You cannot hire your way out of the gap fast enough, and the practices that keep trying will spend the rest of the decade doing the one thing none of them want to do: turning volume away. The more useful way to look at it, she argues on this episode of Flight Path, is as a capacity problem. Not how many radiologists can we add, but how many studies can we move through safely, and how fast can we get the read back to the patient.
That reframe matters because the two halves of the problem are moving in opposite directions. Jessica has watched imaging volume climb for reasons that have very little to do with any single practice. Care keeps shifting to the outpatient setting. The population is older and living longer with more conditions at once, and each of those conditions tends to bring imaging with it. Order sets increasingly default to a scan before a patient is even fully examined. She tells the story of a patient arriving with right lower quadrant pain who gets an ultrasound automatically, sometimes surfacing a critical result before triage is finished. That is imaging doing exactly what it should. For every case like it, though, there are many where the scan was ordered out of habit or caution rather than need. Volume rises either way.
Raleigh Radiology runs 11 imaging centers and partners with 10 hospitals across three states, reading roughly 1.5 million studies a year with more than 40 radiologists, so for Jessica the capacity problem is not theoretical.
Set that against a supply of radiologists that is not keeping pace, and the math gets uncomfortable. So Jessica's team spends its energy where it can actually move the number: capacity. Three levers do most of the work.
Lever one: give the radiologist back their time
The scarcest resource in the building is a radiologist's attention, so the first job is to stop spending it on anything that is not interpretation. At Raleigh Radiology, when an exam is missing information, when notes are incomplete, or when someone needs to talk to a technologist, a non-physician handles it. A credentialing team owns licensure and payer contracting. Assistants take the legwork, and a billing and coding team chases down what might otherwise land back on the radiologist. Strip all of that away and what is left is the work only a radiologist can do.
Some of that time comes back through automation. When a radiologist dictates the body of an exam, the impression at the bottom is often just the top few findings repeated. Jessica's team uses tools that take the radiologist's own dictation and generate that impression for them. Her practice measured the effect at somewhere between fifteen and twenty percent of a radiologist's time returned, which is a large number when the whole goal is reading more studies in a day without adding people.
Lever two: order the worklist so focus holds
A pile of studies is not the same as a plan for reading them. With volume this high, Jessica's team triages, using a mix of human judgment and automation to separate the urgent from the routine so the critical finding gets seen first. They also pay attention to sequence. Jumping from an X-ray to a CT to an MR and back forces a mental reset each time. Grouping similar reads lets a radiologist stay in one mindset and move faster with fewer errors. None of this is glamorous. It is workflow design, and it is where a lot of real capacity hides.
Lever three: make the read easy to reach
The first two levers only pay off if the data is actually available when the radiologist sits down. Fast, reachable imaging is what lets someone read quickly, wherever they happen to be working that day. When the systems are slow, fragile, or stitched together, the time you saved elsewhere leaks right back out.
Which leads to the point Jessica makes most bluntly. Competing for radiologists used to come down to salary and benefits. It does not anymore. A reliable, modern technology stack is now part of how a practice recruits and keeps its readers, because the best of them will not stay somewhere the tools fight them. She quotes a line from a recent conference that has stuck with her: if your technology does not work, you are no longer relevant. The days of a system that works most of the time, unless it is raining or cloudy outside, are over.
The discipline that ties it together
What separates Jessica's approach from a shopping list of tools is the follow-through. Her team does not implement something and walk away. They watch the metrics, track the time savings, and survey the radiologists to confirm a change actually helped. A tool that does not move the numbers does not keep its place. That is a useful standard for any leader weighing the next AI pilot or workflow change: decide up front what you expect it to move, then go back and check.
A quick self-diagnosis
If you want to know whether your own capacity plan is really just a hiring plan wearing a disguise, three questions get you most of the way there. First, how much of your radiologists' time is still going to work that is not interpretation, and who could own that instead? Second, does your worklist actively sort and sequence what gets read, or is it a queue? Third, when you rolled out your last new tool, did you measure whether it saved time, or did you assume it did? Honest answers tend to point straight at the capacity you already have and are not using.
Jessica's read on the next three years is that this only gets more urgent. The shortage is expected to intensify before it eases, and the practices that treat capacity as the problem now are the ones that will still be saying yes to their communities later. The full conversation gets into her path from the modality floor to the CEO chair and why that whole-system view shapes how she leads.
Hear the full conversation with Jessica Bordeaux on Flight Path — available on YouTube, Spotify, and Apple Podcasts.
Keep reading
All articles
Imaging Is the Biggest Data Problem in Healthcare. Here Is the Field Guide.
Imaging is the largest data type in healthcare and the least understood outside the reading room. A vendor-neutral field guide to how it ...
Five Ways to Protect Your Imaging Archive, and Keep Your Data Moving
The five things that most often put an imaging archive at risk, and the simple, preventable habits that keep studies moving and your system ...
The data has to move first: why trustworthy imaging AI starts with access
Trustworthy imaging AI starts with access. Mach7 CEO Teri Thomas on population fit, validating models on your own patients, and why data ...


