Rethink Imaging

EP 23 • July 15, 2025

The Missing Piece in Medical Imaging Safety

Featured Guest

Dr. Madan Rehani, PhD

Director, Global Outreach for Radiation Protection; Professor of Radiology, Harvard Medical School • Massachusetts General Hospital

Listen Now38 min

Also on:

Let's Talk

Beyond dose

Ask us what your imaging data can actually show you.

Start a Conversation

Share This Episode

Subscribe to Rethink Imaging

Most conversations about CT safety stop at the single scan. Dr. Madan Rehani has spent his career arguing that the real question is what happens when scans add up. After more than a decade leading the IAEA’s radiation protection of patients program, work that shaped practice in over 100 countries, he now directs global outreach for radiation protection at Massachusetts General Hospital. In this episode he tells host Chris St. John why current policy amounts to one blinking green light at a busy intersection: full discretion for whoever orders the exam, no yellow light, no checkpoint, no matter how high a patient’s cumulative dose climbs.

The conversation covers the science and the politics. Rehani explains why no national or international body will discount doses received months apart; the only accepted method is to sum them. He proposes a tiered system borrowed from pharmacy, over the counter, prescription, and scheduled, with closer review as organ doses approach 100 mGy. He also names the uncomfortable part: dose-cutting technology already exists, and vendors told him directly that nobody is asking for it. From the 2025 JAMA projection of 100,000 future cancers from CT to advice for an anxious parent, this one stays practical.

Host

Chris St. John

Host, Rethink Imaging • Imalogix

Featured Guest

Dr. Madan Rehani, PhD

Director, Global Outreach for Radiation Protection; Professor of Radiology, Harvard Medical School • Massachusetts General Hospital

Watch the Episode

Key Takeaways

There is no scientific discount for spreading scans over time. Rehani put the question to ICRP leadership directly, and the only accepted method is summing doses across exams. Repair corrections exist for skin reactions in radiotherapy, but not for cancer risk, where even a small fraction of misrepaired mutations can matter.
Current dose policy is one blinking green light: full discretion for the ordering clinician with no advisory thresholds. Rehani proposes a pharmacy-style model, over the counter, prescription, and scheduled, with a yellow light at a few tens of mGy organ dose and closer review near 100 mGy, the level where national and international bodies agree risk is real.
Four of the six manufacturer dose-reduction options from his 2020 European Journal of Medical Physics paper remain unimplemented. Photon counting detectors (roughly 50 percent dose reduction in chest CT at his hospital) and deep learning reconstruction have shipped; more powerful tubes with thicker pre-filters, tube voltage modulation, expanded automatic exposure control, and adaptive bowtie filters have not.
When Rehani asks vendors at RSNA why dose reduction is no longer a flagship selling point, the answer is “nobody is asking us.” Buyer pressure, not technical capability, is the missing ingredient.
Population projections like the 2025 JAMA estimate of roughly 100,000 future cancers from current CT use are system-level planning tools, not personal predictions. Framed with context, they drive smarter policy; framed as headlines, they drive fear.

Full Transcript

Rethink Imaging Podcast Transcript

Guest: Madan Rehani

Host: Chris St. John

CHRIS ST. JOHN 00:01:17 Welcome back to Rethink Imaging. Today, I am joined by Dr. Madan Rehani. Dr. Rehani is a professor at Harvard Medical School and the director of radiation protection outreach at Mass General and Harvard Medical School. Before that, he spent over a decade as the head of radiation protection at the International Atomic Energy Agency, where he helped shape best practices for managing radiation exposure in medical imaging across more than 100 countries.

Today we are digging into a subject that is technically complex, deeply human, and potentially a little controversial, which I am excited about. We are going to talk about cumulative radiation dose. Dr. Rehani, thank you so much for joining us today.

MADAN REHANI 00:02:04 Thank you, Chris, for having me.

CHRIS ST. JOHN 00:02:07 Before we dig in, can you give our listeners background on your career journey? I believe you began as a diagnostic medical physicist, but eventually led the International Atomic Energy Agency Radiation Protection of Patients (IAEA RPAP) program, influencing radiation safety policies. How did that career arc change the way you think about patient radiation protection as compared with the average diagnostic medical physicist in a hospital?

MADAN REHANI 00:02:42 Thank you very much, Chris, for asking this very important question. In short, I will say it changed my outlook immensely. This is a profound and important question with three components.

Number one, as a medical physicist in a hospital, you deal with so many things, and radiation protection is a very small part of it. Whereas in the IAEA Radiation Protection of Patients program, my job was 100% on radiation protection of patients. It is like comparing a primary care physician to a cardiologist, neurologist, or urologist—a generalist versus a specialist. Additionally, I realized that my decisions could affect millions, and I pursued my actions as if I were responsible for patient safety in several billion exams and procedures conducted globally.

Number two, I took ownership of my patient safety program as a father of the program would do for his children—in this case, billions of exams being performed in the world.

Finally, this was a shift from building a CV to building a [unclear: sounds like “LJC”], aiming for results that outlast the tenure. Thank you.

CHRIS ST. JOHN 00:04:19 Thank you so much. That was very succinct; I appreciate that. Working across so many health systems, you have seen everything from national dose registries to paper-based record keeping. In your eyes, what best practice for tracking or managing patient exposure abroad do you think the United States potentially undervalues?

MADAN REHANI 00:04:41 Thank you. Firstly, I admire your question because it shows openness to what exists elsewhere and recognizes that something is being undervalued in the US. Our job at the IAEA typically involves dealing with less developed countries—in my language, we call them Karmabhoomi, the place where all the actions of United Nations organizations go. But at the same time, we learn from best practices in developed countries.

Frankly speaking, when I was in Europe participating in policy planning meetings for the European Commission, I hardly ever got the feeling that radiation protection was viewed as being against the business of radiology. All roads in radiation protection led to making patients safer, and that was it. I wish it could be the same here.

There are many publications showing higher patient doses for similar exams in the US in identical settings. It is like McDonald’s saying that the call for a low-calorie burger based on high risk in a standard burger will hit their business. Consider auto manufacturers historically resisting mandatory airbags and seatbelts in the 1970s and 80s due to concerns over cost and customer [unclear: sounds like “receptors”]. Or consider regulatory culture: the US often places emphasis on warning labels and consumer choices, while Europe leans toward preventive regulation and redesign. The EU bans many cosmetic chemicals and food types that are still allowed with warnings in the US.

Coming back to medical radiation protection, I found that radiation protection is pursued in most countries as a necessity, a common-sense measure, and something helpful for safety. Anything to strengthen radiation protection may evoke reactions based on practicality or economics, but overall not as something that will reduce the volume of imaging. That summarizes the difference.

CHRIS ST. JOHN 00:07:37 In 2022, you participated in an educational debate at AAPM titled “Use of Cumulative Effective Dose.” It was moderated by Dr. Chame [unclear: sounds like “Dr. Chame”], a frequent guest and friend of the podcast. In that debate, you argued that dose tracking isn’t about hitting a single numeric limit like a stop sign. What sort of insights should clinicians draw from a patient’s scan history?

MADAN REHANI 00:08:29 Imagine a traffic intersection with only a blinking green light in a setting with heavy traffic in a big city. What will be the result? Chaos. If the advice is “watch and go,” even with repeated announcements, will it work? Obviously not.

This is exactly the situation we have in radiation protection policy for patient exposure. We only have a blinking green light, giving total discretion to the persons ordering the exams to “watch and go, use your judgment.” Not only that, but there is fear that anything restrictive will stop an ambulance carrying a sick patient. If resistance comes from ambulance drivers, that is understandable, but imagine resistance coming from the traffic police.

I deliberated this issue extensively in my 2021 article in the Journal of Radiological Protection, titled “Old Enemy, New Threat.” You cannot solve today’s problems with yesterday’s solutions.

Another comparison is pharmaceutical drugs, where there are three levels: over-the-counter drugs, prescription medicines, and scheduled drugs. Within scheduled drugs, like narcotics, there is an extensive record of the amount prescribed and taken. Why are we satisfied with only one blinking green light in imaging? It is a primitive approach, and we need to learn from systems that are much more advanced.

CHRIS ST. JOHN 00:11:48 To follow up on the traffic light analogy, what in imaging or radiology would you classify as a yellow light or a red light?

MADAN REHANI 00:12:19 Above 100 mSv of effective dose or 100 mGy of organ dose, there is a wide consensus across national and international organizations that radiation risks exist. One could place these cases into a scheduled category. Somewhere down the line—for instance, 30 to 40 mGy up to 100 mGy of organ dose—one could trigger a yellow light. The specific numbers can be debated, but the structural concept must exist. Currently, the overarching concept remains that the doctor’s decision is supreme and unrestricted, or else the patient will suffer.

CHRIS ST. JOHN 00:13:08 What do you mean by that—that the doctor reigns supreme or else the patients will suffer? Are you talking about patient voices or dose considerations not being part of the conversation?

MADAN REHANI 00:13:25 Some professional societies treat every clinical situation as urgent, as if every passing vehicle were an ambulance. Whenever controlling radiation doses is suggested, there is strong pushback that it will cause patients to refuse needed exams. We encounter resistance to anything beyond the existing framework, which assumes dose should not be a limiting consideration in diagnostic imaging.

CHRIS ST. JOHN 00:114:34 When you mentioned 100 mGy, were you thinking about that as a singular exposure or as a cumulative dose accumulated over time?

MADAN REHANI 00:14:45 There is no scientist from any national or international organization who can provide a correction factor for recurrent exposures, such as five CT scans in a year. I have asked leading scientists, including chairs of the ICRP and ICRP Committee 1, how to calculate dose for patients receiving multiple CT exams throughout the year. They state that the only scientifically supported method is to sum the doses at different intervals.

In radiotherapy, dose effects are adjusted over time because skin reactions undergo cellular repair. However, scientists note that this repair knowledge does not apply the same way to carcinogenic risk. For carcinogenic effects, even if a large fraction of DNA mutations are repaired, the remaining unrepaired or misrepaired fraction can potentially lead to cancer. As a result, no international organization offers a better method than summing doses received at different times. Whether delivered at a single point or divided over time, the cumulative total is what must be evaluated.

CHRIS ST. JOHN 00:16:59 There is an element of fear that comes into play from the patient perspective. As someone relatively new to the field of imaging, I see narratives that create fear about developing cancer from future scans. How do we ensure dose history becomes a tool for better clinical judgment rather than a source of fear-driven avoidance? What safeguards or education should be in place to keep cumulative dose tracking from backfiring or preventing life-saving care?

MADAN REHANI 00:17:58 Cumulative dose history should function as a clinical compass, not a guilt trip. If an electronic medical record flags a drug allergy, it does not mean “do not treat”—it means pause, reassess, and adjust wisely. We need to integrate dose tracking into clinical systems to display interpretable risk tiers rather than raw numbers, much like standard reference ranges in blood tests.

If framed as a patient safety tool rather than a liability metric, dose tracking transforms fear into foresight and empowers clinical teams to make informed, patient-centric decisions instead of defensive ones. However, we must remain open to solutions. Too often, when new approaches are proposed, roadblocks are erected instead of collaborative problem-solving.

CHRIS ST. JOHN 00:19:32 Is it fair to say you agree that we need more research on cumulative dose to establish a more thorough and structured approach?

MADAN REHANI 00:19:44 Yes, which is why companies like Imalogix and others have developed dose management systems. These systems are growing in popularity—initially in Europe, and increasingly in the US. There is a clear need for these systems so that exposure data can be used judiciously.

CHRIS ST. JOHN 00:20:13 May 2025 executive order directed US regulators to re-examine the linear no-threshold (LNT) model that underpins most radiation risk estimates. If regulators weaken or abandon LNT, do you worry that momentum to track and limit cumulative dose will fade, or will clinical culture keep pressing for lower exposure regardless?

MADAN REHANI 00:20:48 My work in recent years focuses on relatively higher patient doses in the tens of mGy or mSv, which will not be affected by these changes.

Do we let faucets run unnecessarily, leave lights on in empty rooms, or prescribe antibiotics inappropriately? Radiation safety follows basic tenets of care. When a medication transitions from prescription to over-the-counter, it does not mean the drug is risk-free; it means the regulatory threshold has shifted, but responsible use remains necessary. Clinical culture must continue driving appropriate utilization with or without LNT. The commitment to minimizing radiation exposure will not fade because dose tracking is about avoiding waste, improving care quality, and respecting the patient.

CHRIS ST. JOHN 00:23:10 How would you explain scientific uncertainty to patients and policymakers without trivializing risk or causing alarm?

MADAN REHANI 00:23:21 Uncertainty is built into nearly everything in medicine. High cholesterol does not guarantee a heart attack, and high blood pressure does not guarantee a stroke tomorrow, but we manage them because the risk increases and we want to prevent harm. Radiation is similar. Low-dose exposure carries a small risk, and while we cannot predict exactly who will be affected, that is no reason to ignore it. When risk is uncertain but avoidable, we act with reasoned caution rather than denial or alarm.

For policymakers, uncertainty is a reason for prudent system design. Just like seatbelts and vaccines, we implement safeguards even when individual absolute risks are small.

CHRIS ST. JOHN 00:24:36 A 2025 JAMA Internal Medicine paper projected lifetime cancer risks from current computed tomography, estimating around 100,000 future cancers from CT usage. That study received widespread coverage in mainstream media outlets like People, NPR, and NBC, often with context or numbers misreported. Do massive population-level projections advance safety conversations, or do they risk causing sensationalism?

MADAN REHANI 00:25:35 Large-scale projections can be useful when framed properly. Public health regularly relies on similar models, such as projecting heart disease from cholesterol, diabetes burden from obesity, or cancer rates from tobacco use. These estimates are designed to help decision-makers plan screening, prevention, and educational programs, not to cause panic.

The same principles apply to CT risk models. When projected responsibly against scan volumes, the goal is to guide smarter policies—such as dose tracking, appropriateness criteria, or alternative imaging modalities. They are tools for system-level insight, not individual predictions. When paired with education that avoids fear and emphasizes context, they advance the safety conversation.

First, as medical physicists, we work primarily on the technical side rather than the clinical front line. Pediatricians directly handle these conversations, though they may consult us for guidance.

I would acknowledge the parent’s concern with empathy. Any parent in that situation would naturally worry. Then I would gently transition to the evidence: CT scans use radiation, and while the risk from a few scans is very small, it is not zero. However, the most critical factor is that those scans were performed for valid clinical reasons to diagnose or monitor a condition.

Having a few CT scans does not mean the child will develop cancer; it means we monitor their care thoughtfully. I would reassure them that we actively look for ways to limit future exposures, utilize safer alternatives when available, apply optimization protocols to perform scans at the lowest achievable dose, and only recommend imaging when it directly benefits the child. Balancing empathy with evidence means addressing the fear without overstating the risk.

CHRIS ST. JOHN 00:27:14 Radiation workers have annual dose caps, but patients do not. Could an advisory dose budget for patients help clinicians think differently about repeat imaging, or could it create confusion?

MADAN REHANI 00:27:52 An advisory dose budget acts like a yellow light: it prompts clinicians to pause and reassess, not to halt necessary care. As cumulative dose builds—especially beyond 100 mGy to a specific organ—applying a framework similar to scheduled drugs is reasonable. We do not prohibit necessary procedures, but we monitor cumulative exposure and require thoughtful oversight. This framework encourages balance, keeping clinicians alert to risk without freezing medical decision-making. Supported by good communication and decision tools, a soft dose budget guides appropriate imaging rather than creating confusion.

CHRIS ST. JOHN 00:28:53 In your editorial titled “CT is Still Not a Low-Dose Imaging Modality,” you argue that manufacturers could already deliver submillisievert levels in CT. What vendor-related technologies or designs are available that have not yet become industry defaults?

MADAN REHANI 00:2930 In a 2020 paper published in Physica Medica (European Journal of Medical Physics) titled “Is it possible to kill the radiation risk issue in computed tomography?”, co-authored with a CT technology expert from Germany, we identified six key technological factors manufacturers could leverage to reduce dose:

More powerful X-ray tubes with thicker pre-filters in filter changers.

Tunable tube voltage settings.

Advanced automatic exposure control incorporating kV and filter thickness modulation alongside existing mA modulation.

Adaptive bowtie filters.

Photon-counting detector technology.

Deep learning reconstruction algorithms.

Of these six, two have been commercialized: photon-counting detectors and deep learning algorithms. In our hospital, photon-counting CT has demonstrated roughly a 50% dose reduction in chest CTs. However, the other four technical capabilities remain largely unimplemented across the industry.

When I ask vendors at conferences like RSNA why dose reduction is no longer their central marketing focus as it was years ago, their response is that buyers are not requesting it. The technical capability exists, but user demand is lacking. When buyers actively demand these features, manufacturers will make them standard defaults.

CHRIS ST. JOHN 00:34:18 It is surprising that demand isn’t driving this, as one would assume a manufacturer offering these dose-reduction features would gain a major competitive advantage.

MADAN REHANI 00:34:30 Looking back at history, in the 1980s, the WHO developed a Basic Radiological System designed for developing countries. It was an X-ray unit that ran on standard electrical outlets or batteries during power outages, with manual arm movements rather than motorized parts. The technical designs were provided to several vendors. However, sales representatives rarely promoted these lower-cost units because they yielded lower profit margins, resulting in minimal commercial adoption beyond donated units. Adoption depends on how aggressively users demand technology and how vendors choose to commercialize it.

CHRIS ST. JOHN 00:36:07 To close our episode today, I have three quick high-level questions for you. Question one: What is one truth in imaging safety that people in the field are reluctant to admit?

MADAN REHANI 00:36:40 That cumulative dose to a larger number of patients is a real safety problem.

CHRIS ST. JOHN 00:36:57 What is one idea in radiology safety that you think the field is getting wrong?

MADAN REHANI 00:37:00 The assumption that patients receiving high doses are few in number and too ill to live long enough to suffer radiation effects.

CHRIS ST. JOHN 00:37:17 If you were in charge of imaging safety in the US, what primary change would you make?

MADAN REHANI 00:37:22 I would aim to virtually eliminate the radiation risk issue in medical imaging so that patients could undergo any clinically necessary imaging exam without safety concerns. It is achievable.

CHRIS ST. JOHN 00:37:37 Dr. Rehani, thank you so much for joining us today on Rethink Imaging. It has been a pleasure having you, and we hope you will join us again in the future.

MADAN REHANI 00:37:54 Thank you, Chris, for having me. I enjoyed discussing these points and hope this conversation generates a positive impact for patient care.

CHRIS ST. JOHN 00:38:20 Thank you. Here at Rethink Imaging, we aim to foster these vital conversations. To our listeners, if you have thoughts or feedback on today’s topic, please reach out to us online or via email. Dr. Rehani, thank you once again.

MADAN REHANI 00:38:37 Thank you, Chris. Thank you to the Rethink Imaging program and to Imalogix for supporting this effort. Goodbye.

Questions from This Episode

Does spacing CT scans out over months or years lower the cumulative risk?
No. Rehani has asked the world’s leading radiation scientists, including ICRP leadership, and the answer is that doses received at different times must simply be added together. Repair corrections exist in radiotherapy for skin reactions, but for carcinogenic effects even a small fraction of unrepaired or misrepaired mutations can lead to cancer, so no correction factor applies.
Should patients have dose limits the way radiation workers do?
Rehani argues for advisory tiers, not hard caps. A soft yellow light at a few tens of mGy organ dose would prompt clinicians to pause and reassess rather than stop. As cumulative organ dose approaches 100 mGy, he suggests a scheduled-drug model: imaging stays available, but its use is monitored and reviewed.
If low-dose CT technology already exists, why isn't it the industry default?
Demand. His 2020 paper identified six manufacturer-side dose reduction options, and only photon counting detectors and deep learning reconstruction have reached market. Vendors at RSNA told him plainly that nobody is asking for dose reduction anymore, and he says the moment buyers ask, manufacturers will deliver.