Rethink Imaging

EP 16 • April 3, 2025

‘The Real Harm Is in Missing the Diagnosis’ with Dr. Ehsan Samei

Featured Guest

Dr. Ehsan Samei, PhD

Chief Imaging Physicist, Duke University Health System; Reed and Martha Rice Distinguished Professor of Radiology • Duke University Health System

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 medical imaging start with a tradeoff: the scan finds the problem, the radiation carries a risk, and the two sit on opposite sides of a scale. Dr. Ehsan Samei has spent his career showing why that framing fails patients. In Part 1 of this two-part conversation, the Chief Imaging Physicist for Duke University Health System walks through what the evidence actually says about radiation exposure, why population-based dose thresholds make poor stand-ins for clinical judgment, and where the Linear No-Threshold model gets misapplied in everyday practice. His argument is blunt: when imaging is used appropriately, the benefit outweighs the risk by orders of magnitude.

The harm Samei worries about is the scan that never happens. He points to evidence that suboptimal CT imaging, or skipping the exam altogether, carries a risk about five times higher than any potential risk from the radiation itself. That reality reframes the whole safety conversation. He also argues for patient autonomy: it is the patient’s body, the patient’s diagnosis, and the patient’s risk, so imaging decisions belong in a shared conversation between the expert who knows the procedure and the person receiving it. The episode closes with key takeaways and a preview of Part 2, where the discussion turns to whether the field’s focus on dose has gone too far.

Host

Chris St. John

Host, Rethink Imaging • Imalogix

Featured Guest

Dr. Ehsan Samei, PhD

Chief Imaging Physicist, Duke University Health System; Reed and Martha Rice Distinguished Professor of Radiology • Duke University Health System

Watch the Episode

Key Takeaways

Samei calls imaging “a gift” for sorting out what is happening inside the body, and rejects the standard benefit vs risk framing because it treats two wildly unequal quantities as peers. In his words, the benefit outweighs the risk “by orders of magnitude.”
The Linear No-Threshold model is widely cited and widely misunderstood in clinical contexts. Samei walks through what high-dose and low-dose exposure data actually show, and why population-based thresholds are a poor substitute for clinical guidance on an individual patient.
The risk tied to suboptimal CT imaging, or not doing the CT at all, is about five times higher than any potential risk from the radiation. Delayed or denied imaging is where the real harm lives.
On cancer attribution: roughly 40 percent of people will get cancer in their lifetime. If imaging adds 0.01 percent to that, there is no way to pin a given cancer on a scan rather than any other exposure, which is why fear-based risk framing falls apart.
Patient autonomy belongs at the center. It is the patient’s body, diagnosis, and risk, so imaging should be a shared decision between the expert who knows the procedure and the patient who lives with the outcome, rather than a paternalistic call made for them.

Full Transcript

[00:00:01] Dr Ehsan Samei: Briefly look at them. I have already rehearsed in terms of

[00:00:05] Christ St John: Mean honestly, I think that’s for the best, because I think I’m going to probably deviate a little bit. I’ve been finding more and more, I use my prepared questions less and less. I I got to record with Nina Cutler yesterday from Stanford who I’ve been hassling for a little while. And I think I used my opening question, and then that was it. After that, we just talked. Um, now it’s like, this is perfect. Alright. Uh, you ready?

[00:00:44] Dr Ehsan Samei: Yes. I think so. Let me make sure I put your face where my camera is so I can focus on you and not myself. Okay. I’m good.

[00:00:54] Christ St John: Alright. Beautiful. Um, let me actually do that. Let me rearrange this a tiny bit so I look. There we go. Beautiful. Okay. Well, uh, hi, everybody. Welcome back to Rethink Imaging. Uh, I am so thrilled to have, uh, probably the bet like, the closest friend of the podcast that we have, uh, doctor Hassan Soumay, um, podcast regular, uh, medical physics legend, uh, and somebody who I think I can call a friend at this point. Uh, doctor Soumay, welcome back again.

[00:01:28] Dr Ehsan Samei: Thank you, Chris.

[00:01:31] Christ St John: Um, so I’m I’m super excited to have you back because, uh, we’ve been doing this show for two years now, which is crazy to me. I believe you were our second guest ever. And since then, we’ve gone on a lot of different journeys. We’ve talked way too much about those. Um, we’ve talked a significant amount about AI and imaging. We’ve been we’ve been all over the place talking about technologies and people and roles and all of this different stuff. But I wanna kind of get back to basics a little bit and and get back to the core of medical physics because I feel like we’ve gotten so inside baseball and we’ve gotten so granular with some of these conversations that we’ve kind of been drifting from, like, the core competency of of medical physics role in imaging. Um, and I feel like you are the the perfect person to have that, uh, that conversation with me. Um, so I just super high level. Right? When you think about medical physics role in a clinical health care setting, what what is the role of the physicist? What is the role of physics? Uh, and and where does it kind of fit in this organizational map, at least today?

[00:02:58] Dr Ehsan Samei: Mhmm. Um, I guess the the answer lies in the very term itself. Medical physics is physics and medicine meeting one another. In the context of radiologic medicine, uh, which is the primary focus that we have been dealing with here, is radiology and physics meet one another. Uh, many of us think of medical physics as physics of medicine or physics in medicine, um, or physics of medicine, I’d like to think of and that’s that’s definitely appropriate depiction of what medical physics is. Is the physics related to medicine or radiology in this case and physics as it’s practiced in medicine? But, uh, increasingly, we have been moving towards what I call physics for medicine. In other word, medicine is the ultimate calling that we have in this space. The question that I ask myself and physicists in the field should ask themselves is that, what is it that medicine needs from physics to make it better? So the way one way to think about it is that physicists, by their training, by their competence, they are quantitative scientists. We spend our time going to graduate school, getting doctorates and masters and so on, trying to understand data, try to interpret data. We plot data. We look at the trends. We look at predictions. We have principles. We do experiments. We look at the results that we get from the experiments, quantitative results from the experiments to see how those results, uh, verify or invalidate the assumptions that we had about the area that we practiced. So this is this is a wiring that medical physicists have. So in the modern practice of medicine, I think that wiring can be useful. It gives a quantitative edge and analytical approach towards understanding the the medical processes. Science is very quantitative in nature, so we are individuals who add quantification, analytical mindset, methodological approaches towards the practice of medicine.

[00:05:43] Christ St John: Okay. Uh, I’ll I I can I can I can run with that right now? And so when you are thinking about, like, a a physicist who is employed, um, you know, at at some sort of health care facility. Mhmm. I actually, you know, let me let me let me reframe that a little bit. Uh, editors, you can scratch that little bit out. Um, one of the things that I feel like I have noticed in my last two years, uh, working in imaging tech is there seems I I I hear from a lot of physicists, and I talk to a lot of physicists. And very often, there is a little bit of a sense, and I wanna I wanna watch my wording, but there there there can be a sense that radiologists um, and physicists, like, the channels of communication between them are not always open and flowing in the ways that maybe the system was designed to or that we would hope. Um, in fact, I, um, I I I’d rather not say who it is, but I recently spoke to a physicist, uh, Hassan, who I think you know pretty pretty well, who recently changed where they’re working. And one of the key things that they said about their new role is, man, the radiologists are just so involved, and they really, really care. And they’re coming to me about the images, and we’re talking through things, and we’re working through things at a level that I haven’t seen before, which to me sounds, you know, beautiful and like an ideal, but it is absolutely not, um, the norm. And I’m I’m I’m curious from your perspective, why do you think that that perhaps isn’t quite so so normalized in practice?

[00:07:41] Dr Ehsan Samei: Yeah. Um, multiple reasons. Um, so one of the hallmark of a civilized workforce or society is that we use experts for for different areas of needs. I never ask electrician to come into the plumbing in my house or vice versa. Right? So I would say that radiologists have very specialized skills in terms of understanding images, in in terms of putting the images that they’re getting in the context of broader patient care. They look at the patient chart. They have a lot of information. They lot of, uh, deep insight into signatures of the disease and what is normal, what is abnormal. That’s their space that they can fully claim as their space. For us as physicists, our job is to understand how anatomical and physiological properties of the body can be best made visualized, how can we extract that information in most effective, efficient, cost effective, dose effective fashions. That is our wiring. This is our space. So I think the expertise of radiologists and physicists are very complementary. In the ideal world, they would work with one another. Of course, they they are touchpoint. We both in the context of medical imaging, we both work on medical imaging, but radiologists are more bending toward interpretation, and we are more bending toward technological aspects of the image acquisitions and image formation process for the care of the individual patient. That’s that’s just how we sort of chart the territory of radiological, like, uh, practice. Of course, we are not the only two. There are other people involved, the stakeholders. There are acknowledges who actually, you know, engage with the patients and image the patients and so on. Right? Uh, so if we respect each other’s expertise and work together, we can synergize our expertise toward the towards doing right by the patient, which is ultimately the the the goal of medicine anyway for a given patient in front of us. We need to sort of work together, you know, to deliver the best of care for the patient. Now why is it that in some places and many places, perhaps it doesn’t work very well? Um, one area is that we sort of commodify the technology. We feel like, why do you actually need to have a physicist? Right? I can’t just go to the vendor, buy the machine, bring the patient to the room, ask the technologist to acquire the images, and they just read the cases. What’s wrong with that model? And many places actually perhaps use that model. Um, I would say the problem with that is that we forget about the fact that imaging process has I’m using physics jargon here. There are too many degrees of freedom. In the word another way of saying, there are too many things you can adjust, and those adjustments change the output of the images that you acquire. Images are never perfect reflection of the reality. Never. Have never been. Right? That’s one of the reasons, for example, we have, uh, complementary modalities. Why do you just do an MR and not a CT or a CT and MR? Because they two they give you two different dimensions of reality of the patient. So each each dimension itself is not enough. And depending on what dose you use, what setting you use, what reconstruction you use, what, uh, how you frame the images, uh, field of view, pitch, all of that those factors affect the images that you’re getting. And on top of that, patients come with different makes and models. We are different. My needs are different from your need. Your context is different from my context. Right? So in this sea of degrees of freedom, how do I, um, uh, appropriate the right setting for that patient? I think that’s a very complicated, um, question to answer.

[00:12:04] Christ St John: Mhmm.

[00:12:04] Dr Ehsan Samei: I think the answer requires the complimentary expertise of a radiologist and the physicist and the technologists. If you think any of us think we got it, I think we have not gotten it. So in order to get this thing to work, we need to have humility in practice of our expertise that we don’t have it all. We need each other. Uh, we need to have respect for one another because we don’t have it all. And why do we need to work together anyway? Because at the end of this world, there is a patient who’s supposed to benefit from our, you know, elevated expertise that we claim.

[00:12:47] Christ St John: Yeah. I mean and you’ve, I mean, you’ve used the phrase. Right? What is it? Images about the patient, not the machine? Or imaging is about the patient, not the machine? Right? I mean, and and what does that actually mean for what a radiologist is seeing on the screen and what a technologist is doing at the council and what you or physicists are doing behind the scenes monitoring protocols or adjusting protocols? Like, what can you expand on that

[00:13:14] Dr Ehsan Samei: a little bit? Yeah. Yeah. So, I mean, it goes back back to what the phrase that I just said a little bit earlier that the images are never a perfect rendition of reality of the patient. Mhmm. Um, is is a depiction of it. It corresponds to that reality. It’s highly relevant to care care of the patient, but it’s never perfect. And it’s malleable. It’s changeable. It’s fluid. You could think that’s that’s that’s another way of saying that there are different degrees of freedom. So image is a fluid construct that we capture from the patient based on how the images are set up. Right? Um, and we wanna do right by the patient. Practice of medicine, though, often and frequently, not done in a patient specific manner. It’s done in an aggregate manner. Mhmm. Let me give you an example. Uh, when you go to a patient and we give you to a go to a physician, and, uh, they wanna do some procedure, and they say there is a five percent of chance of this and fifth 14% chance of that and twenty percent chance of this, uh, and I’m suggesting you should do, you know, course b as opposed to course a or c. Right? Uh, these are typical things that you hear from a from a physician when you when you go, uh, what does it mean fourteen percent? That mean I only affected fourteen percent of me get affected by it? No. That means there’s a population out there, a hypothetical population, that fourteen percent of that population would be inflicted by this side effect that I’m talking about. So, essentially, I’m not seeing you. I’m seeing you as an aggregate, as a representative of a population. And there is nothing wrong with that, uh, because that’s how we practice medicine. We always average. We always look at the mean value. We look at the confidence intervals. So our care is never about the individual patient. It’s it comes from the aggregate data. Our our confidence our science come from aggregate data, but that aggregate data needs to be applied to the individual patient. That individualization process is something that can be significantly helped with with, um, understanding the nuances of the image acquisitions and understanding nuances of the patients. And that’s the personal decision care can can meet the reality. See, think about it this way again. I said that earlier. Ultimately, we wanna do right by the patient, not by an average patient. There is no such a thing as an average patient, but there’s an ex I have not met one.

[00:16:00] Christ St John: So, I mean, how how do you how do you even begin to think about approaching that? Right? I mean, I think it is it is it is very safe to say that, um, in in a lot of locations across the country, a lot of folks are referring to, uh, you know, like, manufacturer suggested protocols. Right? Mhmm. They’re they’re kind of following following the guidebook that they were given. Um, but because of that, those it’s not like competing vendors are constantly talking to each other and comparing their doses and comparing their image quality and sharing all of that with each other and trying you know, it it it’s not like the manufacturers are are always trying to do that. And and because of that, we get a we get a lot of, uh, variability across different technologies. Um, I’m I’m kind of curious. How much of what happens at the console is really being driven by the vendor, by the physicist, and, like, how do we reframe it towards the patient?

[00:17:01] Dr Ehsan Samei: Yeah. Uh, I think manufacturers, rightly so, um, would suggest protocols to be used for, um, imaging patients. And you would imagine that’s the right thing to do because, ultimately, they make the machine. They understand the machine the best. But those those choices are often, uh, are informed primarily by the, um, attributes, the specification of the machine itself, uh, but also by this aggregate feedback that manufacturers have received from from practices out there. Right? They they do their best, and they’re doing an ethical job, and there is nothing wrong with that process at all.

[00:17:47] Christ St John: Oh, yeah. And I’m not I’m not trying to poo poo Right. Manufacturers. Sorry, guys.

[00:17:51] Dr Ehsan Samei: It’s actually great. But but, ultimately, manufacturers, by the virtue of being manufacturers, they’re oriented toward the products themselves, uh, while medicine is about the patient. So another way to think about it is that job of a physicist or job of a clinical institution in general is to make sure the diversity of the products that are available at that facility are somewhat commodified, is a negative term, but but but perhaps makes sense here. Commodify, so it would reflect what the patient needs for the care of that individual patient. I do not care. I should not care as much whether I’m using a Siemens scanner or a Canon scanner or a Genius scanner or a Phillips scanner. I regardless of, you know, what scanner I use, I the scanner needs to do what the patient needs. In fact, manufacturers will also say that. Right? Now if you show me a picture of your amazing vacation that you had in Hawaii, my first question should not be, oh, what camera did you use? That camera needs to be transparent. Right? Uh, ultimately, it needs to be about the patient. So we need to, and because I, uh, there are too many degrees of freedom as I’ve mentioned, those degrees of freedom complicate the process because different devices have different degrees of freedom. Right? So my job I I remember when I started at Duke, there were there were two manufacturers’ devices that we had here, and there were radiologists that were in Camp A or Camp B. They said we we hate the images from the scanners in Camp B, and some people hated the image of our Camp A. And I just didn’t understand why that is. I said, why don’t you adjust the settings of the two machines so they would be similar to one another? And in fact, that became one of the first projects that we did at Duke, uh, to adjust the manufacturer’s settings with the with the constraints that are available to us to make the images according to what the image is supposed to provide. If I need to see the liver lesion, if I need to measure the size of the long term, if I need to measure the stenosis in the cardiac disease precisely, how can I adjust the settings that are available on two different machines? And the adjustments are called differently, by the way, so I can deliver that outcome. So the radiologist would look at the patient, not to images from the manufacturer a or b. That is an essence of patient specificity. That’s the essence of patient centric care. That’s the specific role of a physicist in this space. And so I think we would start I think we should start with the what manufacturers recommend and adjust so they can meet that ultimate goal. I would add one more thing here. I would and why the protocols are so variable across the space. So one is the, uh, of course, different makes and models of machines. Second thing is that different manufacturers recommend different things because they prioritize different things. One manufacturer might prioritize, for example, temporal response to be precise, while the other one might prefer a spectral response to be more precise or vice vice versa. But there is also a degree of, uh, agency of the users of the machine. So we like to fuss around with the options that are available to us. Otherwise, we don’t own it. So nobody likes to be told this is the machine, and that’s the way you’re supposed to use it. Maybe some people would be happy to do that. But most of our intelligent, uh, you know, thoughtful people, either physicists or radiologists or technologists, want to fuss things around fuss around because they think they can make it better. And that’s also a good thing because sometimes there are amazing things that can come about that we have not foreseen, manufacturers have not foreseen, radiologists have not foreseen, and so on. So that level of customization or accommodation per personal preference is also within the landscape. And that would that’s why protocols at Mayo Clinic is different from protocols at Duke and Wisconsin, each one of us, even for the same machine. Um, and I would say even on top of that, sometimes you might wanna go out of that guidance that even you have developed because that particular patient has a particular condition that you really wanna make sure that you do things right. For example, this is a liver transplant patient who’s coming back with repeated infections. So that is a context of this particular patient care. Now you tell me, should I use the typical liver protocol that the manufacturer recommends, or should I make certain accommodation for that specific patient? I think that accommodation is appropriate as as long as I know what I’m doing. That would add some degree of variability in the way we practice medicine. But that variability is appropriate because it’s done for the personalization of the care for that individual patient. We are doing right by the patient. So in the process that we are trying to quantify medicine and reduce the variability, I would argue that some variability is bad, that you for example, for the same patients or patients that have similar attributes, we use different protocols. We need to minimize that. But certain variabilities are actually good because we are accommodating the patient according to what the patient needs. So another way of saying some degree of variable in the practice of medicine in the way that we set up the protocols is called for. The job of us scientists, as physicists, as radiologists in this space to sort out what is good variability, what is bad variability, what is the right protocol, what is the wrong protocol in this space where we have many degrees of freedom that things can go wrong? So we can appropriate the resources at our disposal for the betterment, for the best care of the individual patient in front of us, not the average patient in front of us.

[00:24:39] Christ St John: Yeah. So I wanna I I wanna touch on this this this side of the conversation, which has, like, a a layer of idealism baked into it to some degree. Right? Yes. Yeah. And and and not being dismissive. Right? But, like, you know, I’ve I’ve I’ve listened to you speak tons over over the years. Um, and I’m I’m thinking back to this slide that you have about, like, all of the different qualities that apply to different patients. Right? It’s like it it was a big number. It was like 50 odd signifiers or something. Do you do you remember what that number was off the top of your head? If not, it’s

[00:25:16] Dr Ehsan Samei: not I don’t remember. Right.

[00:25:17] Christ St John: Right. But it’s a lot. Right? You know? Um, race and size and height and different aspects of medical history and

[00:25:25] Dr Ehsan Samei: Right. Right.

[00:25:25] Christ St John: Yes. Mhmm. Um, right. Just like all of the you know, it’s it’s a seemingly unending list, uh, to use a little bit of dramatic flare in my language, um, of all of these things that make all of us different. And if you answer every question on the list, you don’t have an average patient. You have this patient, and you have all the information that you would need to image them ideally. Right? So so in this ideal structure, I’m curious, like, what the organizational workflow would look like. And then and then after we touch on that, I wanna I wanna kinda pivot to the other side of it, talking about, like, capacity and current day volumes and and kind of approach things from the other side as well.

[00:26:07] Dr Ehsan Samei: Yeah. Yeah. Yeah. Yeah. Uh, so the the the the slide that you were talking about is a depiction of trying to recognize the the that that the context of care for different patients are different. So, again, if the screening exam is a pediatric exam, is it is it follow-up exam, is it chronic exam, is it transplant exam, each one is a different context. Right? If the patient was a previous smoker or not smoker matters. If this is something that patient is really sensitized and make sure any additional growth is not missed, while the other patient’s saying, actually, I’m afraid of radiation. I really wanna make sure that you use the minimum radiation possible. So all of these are related to context of care, all of that. And, by the way, patient’s voice is also significantly important in the space. I don’t wanna ignore that.

[00:27:03] Christ St John: Of course.

[00:27:03] Dr Ehsan Samei: I don’t after all, it’s if I’m imaging you, Chris, this is your body.

[00:27:07] Christ St John: Hell, yes.

[00:27:10] Dr Ehsan Samei: You you know, I’ll have to listen to you. If I, like, if I’m the patient, I have to listen to me. So you can’t say, well, we’re gonna ignore you because this is the right thing to do. You can’t do that.

[00:27:21] Christ St John: And I’m the I once again, I’m also the worst imaging patient because I want that dose up. Right. Jokes aside, though. I’ll I’ll stop

[00:27:30] Dr Ehsan Samei: it. I’m actually in the same zone, by the way, but I might have used an example in the past. But, anyway, so you could This is now a physicist talking about this, see how the physicist approaches. So a nonphysicist will say, well, the patient’s stuff matter, and you have to adjust things according to patient’s needs. Right? For that, what I just said, you don’t need to have a physicist. You just can do adjustment as you go. But you don’t wanna be also anarchy in the space. There is a way to organize a space. So we said, okay. How about we identify, like, 150 to 200 labels or tags that we can associate with each case? So now I am putting the attributes of the patients as I understand it from the data that is coming for the medical record and put the patient, uh, make the patient associated with a series of tags. Now if I know the tags for a given set of tags, then I can aggregate different patients together in a smarter way. So I can, for example, put all the critical care patients on transplant list for kidney into a sub category of its own and define what would be most appropriate for that subgroup as opposed to bundle all of them together and say abdominal CT. You see? That’s a way of essentially making that personalization, accommodation, contextualization in a systematic way. This is a good example of how a physicist meets a medical need. Right? Something that was, um, subjective, somewhat even we can say arbitrary, is being, uh, quantified, is being categorized, uh, is being streamlined through a systematic process. But, um, that is the that is I I forgot about forget what the original question was. I think I went to No. No. That’s Different.

[00:29:44] Christ St John: I mean, that it was a great answer because I I I gotcha. So, I mean, thinking thinking about that in practice. Right? Like, all of like, management of all of these tags and, uh, variability in imaging as a result of those tags feels like a gargantuan project. Right? Even even just where things are now in imaging. Right? Like, we still have all of this massive variability. You know, we have we have what we have in terms of we have some diagnostic reference levels, little little tag for, uh, the new radiology publication, um, published by Kunal et al. Uh, doctor Sime is on that publication as well Yeah. For for coming out with new DRLs and, like, giving giving facilities the the opportunity to see what everyone else is up to. But just thinking about the scale of what you’re talking about is is massively intimidating. Right? Adding adding all of these tags to an already rich and complicated imaging process? Like, how how do you consolidate that from your perspective?

[00:30:56] Dr Ehsan Samei: Yeah. I mean, that’s I feel like, by the way, I don’t think it’s super complicated because I think we have the data already at our disposal. And the data is I mean, electronic medical record essentially is a rich space that data is present in that space. And AI and sorting out, uh, resources into different buckets is a process that is can easily be done. Um, I don’t think it’s it’s super complicated. I I’m still advocating for DRL and things like that what that are based on larger, what I would like to call macroaggregates. Mhmm. What I’m advocating for is microaggregates. Microaggregates based on sort of smaller tax. Right? Right. Not small attack. A smaller aggregation based on the tax. I think we need to start the process of having the more generalized, um, description as we have right now and gradually over time, stratify the data in a more in a refined in a in a more refined fashion. That’s that’s all it is. Um, meanwhile, because we we can never we are not there yet, and we need to accommodate the patient cases as they come, I think we need to, um, think of what we do in terms of guidance and not the law. So you could think about the practice of medicine can be based on sort of more of a legal approach saying, this you shall do, this you shall not do. Guidance saying, we recommend you do it this way, but use your own judgment the to the best that is for the patient or in a super personalized personalized fashion that, you know, this is a specific thing that needs to be done for that patient. I think the third one is gonna be difficult to achieve at the moment because the system that I just we just scrubbed is not there yet. But I don’t think we should we should move toward the sort of middle ground that our medicine becomes more guidance based and that’s legal based. So I personally, for example, do not like the idea of thresholds. Anything above a threshold is bad. Anything below a threshold is good. Because, first of all, thresholds tend to be monodimensional. Like, those above this bad, those below this is good. Give you a sense of confidence that, um, if you are below the limit, you have done right by the patient, might not be so because you’re small dimensional. It’s like all the shirts that are less than $40 are good. All the shirts that are, you know, more expensive than $40 are bad. It’s like, you you can’t do that. Right? Depends on what you’re getting. Right? Yeah. Uh, that’s that’s example of a mono dimensional approach. Right? It’s a multidimensional approach and needs to be guidance based. And we I don’t think what I’m saying is very revolutionary. We are already practicing that sort of human judgment everywhere. Right? We accommodate people. We we adjust. We so I think we need to allot that adjustment. That does not mean we need to throw away all the, you know, legal recommendations out there, but we need to give people a little more agency, a respect that they are experts and they knew what they know what they do and they are trying to do right by the patient. Provide professional guidance, but provide also flex flexibility along with it.

Questions from This Episode

Why does Dr. Samei say the benefit vs risk debate misses the point?
Because the framing implies the two sides are comparable, and they are not. Samei argues the benefit of appropriate imaging outweighs the risk by orders of magnitude, so weighing them as equals distorts the decision. Minimizing risk still matters, but the benefit has to stay central to the process.
What is the Linear No-Threshold model and why does it cause confusion?
It is the model that assumes radiation risk scales down linearly with dose, with no safe floor. Samei explains that it is frequently misunderstood when applied to clinical decisions, and he separates what the data show at high exposures from what can actually be claimed at the low doses used in medical imaging.
Can avoiding a scan be riskier than the radiation itself?
Yes, and by a wide margin. Samei cites evidence that suboptimal CT imaging or skipping the exam carries a risk about five times higher than any potential radiation risk. A missed or delayed diagnosis is a concrete harm, while the incremental radiation risk is small and often unmeasurable.