Rethink Imaging

EP 15 • March 20, 2025

Radiation Optimization: Dr. Kanal’s Legacy with the Dose Index Registry

Featured Guest

Kalpana Kanal, PhD

Professor and Director, Diagnostic Physics Section • University of Washington, Department of Radiology

Listen Now35 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

Dr. Kalpana Kanal, PhD, joins host Chris St. John to trace the rise of the ACR CT Dose Index Registry, the national program she helped lead as vice chair and then chair. She explains how the registry lets any facility, from a small community hospital to a major academic center, submit de-identified CT dose data and compare itself against peers across the country. Kanal walks through the two landmark papers built on registry data: national diagnostic reference levels drawn from roughly 1.3 million adult exams, and a pediatric follow-up that benchmarked doses by both age and size. She also covers the practical hurdles, from protocol mapping to the limits of DICOM header data.

The conversation then turns to what came next. Since becoming the American Board of Radiology’s Diagnostic Medical Physics trustee in 2017, Kanal has built a forensic CT program at the University of Washington, scanning more than 150 cadavers in partnership with the medical examiner’s office. In 45 to 50 percent of those cases, CT answered the cause-of-death question without an autopsy. She makes the case that optimization, not minimization, should drive dose management, shows how small facilities can set their own reference levels with a spreadsheet and 20 to 50 patients, and previews where the coming CMS outpatient quality measure fits in.

Host

Chris St. John

Host, Rethink Imaging • Imalogix

Featured Guest

Kalpana Kanal, PhD

Professor and Director, Diagnostic Physics Section • University of Washington, Department of Radiology

Watch the Episode

Key Takeaways

The Dose Index Registry is an affordable entry point into dose monitoring: facilities submit de-identified CT dose data to the ACR and compare themselves against similar hospitals nationwide through what is now a live interactive tool.
The adult diagnostic reference level paper drew on roughly 1.3 million exams to set national benchmarks for the top 10 CT protocols; the pediatric paper went further by benchmarking doses as a function of both age and size. DRLs sit at the 75th percentile of the data, achievable dose at the 50th.
Protocol mapping is the hardest part of DIR participation: a head CT at the University of Washington has to map to common terminology so it compares apples to apples with a head CT at MGH. The registry only sees what is in the DICOM header, and it never sees image quality.
Kanal warns the field may be overcorrecting toward low dose. The goal is an optimized facility, not the lowest number: best achievable image quality at the minimum dose the radiologist can confidently read.
Her forensic CT program at UW has scanned more than 150 cadavers, and in 45 to 50 percent of cases the CT answered the cause-of-death question without an autopsy. The medical examiner’s office is now installing its own CT scanner.

Full Transcript

Kalpana Kanal Mastered Audio Transcript

[00:00:00] Chris St John: Welcome to Frame by Frame, Rethink Imaging, a podcast by Immalogics. Here, we explore the intricate world of medical imaging, aiming to dissect the field and inspire both professionals and curious minds alike. I’m your host, Chris St. John. Today on Frame by Frame, Rethink Imaging, we are thrilled to welcome Dr.

[00:00:23] Kalpana Kanal, a renowned expert in diagnostic medical physics. Dr. Canale has made significant contributions to the field of medical physics, especially in CT. She was involved first as vice chair and then as chair of the American College of Radiology’s CT Dose Index Registry, a key initiative in tracking and optimizing radiation exposure.

[00:00:46] After she became the American Board of Radiology’s Diagnostic Medical Physics trustee in 2017, she stepped back from the DIR. Dr. Canal has continued to focus her research particularly in areas of radiation safety and [00: 01:00] imaging advancements, and recently established an imaging physics residency program at her institution.

[00:01:05] Today we’ll discuss her involvement with the DIR. And her current research focus, as well as the future of medical imaging. Welcome, Dr. Kunal.

[00:01:14] Kalpana Kanal: Thank you, Chris. Nice to be here.

[00:01:16] Chris St John: It’s so good to have you here. Dr. Kunal, your career has spanned a wide range of research and clinical areas within medical physics, especially related to radiation safety and imaging.

[00:01:28] What initially drew you to this field, and how did your journey evolve to focus on diagnostic imaging?

[00:01:35] Kalpana Kanal: Yeah, that’s a great question, Chris. So I came to the U. S. in 1989 to do a master’s in pure physics. And as I was getting done with that, I realized that my personality is not suited for sitting in a lab and doing, you know, bench research.

[00:01:50] I’m more a people’s person and I was started trying to look for other opportunities that can use my background. education in [00: 02:00] physics. And someone told me about a radiological sciences graduate program, PhD program at UT Health Science Center, San Antonio. And so I checked that out and I decided, oh, this looks really interesting with the medical aspect of physics.

[00:02:14] And that’s how I got involved and enrolled in the PhD program over there. While I was there, I was really interested in becoming a therapy physicist, but my research project for my Ph. D. was actually with the diagnostic imaging physicist, Dr. Gary Fullerton at UT Health Science Center, San Antonio. And that’s where my journey with diagnostic imaging started, because I spent all my research on my Ph.

[00:02:36] D. in that field, and then I got both certified in diagnostic medical physics. So that’s led me to the where I am today.

[00:02:46] Chris St John: Amazing. And so not to dive straight into the dose index registry, but we’re going to dive straight into the dose index registry. So you were the vice chair and then chair of the dose index registry.

[00:02:57] For those listeners who are less [00: 03:00] familiar, a pivotal tool for tracking and optimizing radiation doses. How did you get involved with the registry and what was your role?

[00:03:09] Kalpana Kanal: Really good question. Again, as any young, early career physicist, you know, we look up to our mentors and our senior colleagues to get involved in committees and organizations.

[00:03:21] Dr. Rick Morin, who’s a very well known medical physicist, retired now, actually invited me to be part of the Dose Index Registry. So I started out with him being chair and part of the committee and, you know, shadowed him in that committee. And when he rotated off. He asked if I would become the chair of that committee.

[00:03:39] So that’s how I got involved. I was invited by another very well known medical physicist, and that’s where my journey began in the ACR, Dose Index Registry.

[00:03:49] Chris St John: Yeah, and can you just give, like, a little bit of a background on, should I use the word program? How should we even classify it?

[00:03:56] Kalpana Kanal: Yeah, it is the American College of Radiology Dose [00:04:00] Index Registry program.

[00:04:02] You could call it program. Absolutely. It’s for CT. And what it is, is basically, it’s actually a really cool thing. They introduced that, if I remember correctly, Somewhere around, oh boy, 2004 or 5, I probably should have looked it up, but basically the registry came out and it was a way for hospitals, small community hospitals, private practices to enroll in it and send your doses to the ACR with all your PHI information taken out and where you had the opportunity to look how you did with respect to your peers.

[00:04:38] So it’s really a registry of doses. From various organizations, small hospitals, big hospitals. And you classify what type of hospital you are and you can compare your doses to that other hospital, similar types of hospital and see, are you in the same range of doses as your peers? And are you too high?

[00:04:54] Are you too low? Or are you fine? And I think it’s a really a good way to assess that, uh, with an [00: 05:00] outside organization, right? So you’re sending your data and it was relatively financially feasible, not very expensive. And that’s how I got involved too. We became one of the first sites, we were part of the first sites.

[00:05:12] That as soon as ACR was introduced, we enrolled our program to send our doses to the, uh, Dose Index Registry.

[00:05:19] Chris St John: Very nice. Yeah, you know, it’s funny, I was, I can’t remember who I was talking to recently, but I was talking to somebody, and they were like, oh, there aren’t, like, fully standardized doses for all CT procedures?

[00:05:31] And, like, no, definitely not, but it’s, like, it is so cool to, you know, see the origins of, you know, people sharing all of this information, Progressing. And so, given your extensive experience with the Dose Index Registry, or I’m just going to say DIR for the rest of this, what do you think has been, you know, the impact of the Registry on dose management in clinical practice?

[00:05:56] Kalpana Kanal: Oh, I think it’s been significant. As I said, that any [00:06:00] hospital or small community hospital, private practice DIR. It’s not a very big financial burden. It’s quite affordable. And if you have not been doing those management at your facility, this is one way to get started and see where you stand, you know, it gives you reports anytime now.

[00:06:18] I mean, when I started, it was not a live interactive tool. But now, if you go on online, it’s a live interactive tool. You can look at your doses at any time, you can compare it to other facilities. I think it’s a great tool to have at your disposal if you do not have a full, really expensive dose management software available at your hospital.

[00:06:39] The other thing is that’s different about the DIR, it lets you compare your doses with facilities across the country, unlike your own management tool just only looks at your facility. So I think it’s a really good, useful tool to participate in and see how you’re doing. I think it has had a big impact, especially in the United States, because there was nothing like that before DIR.

[00:06:59] [00: 07:00] So having some resource. To collect data from across the country and then establish guidelines and establish landmarks or benchmarks. for doses were great, would be great, right? So the impact was that based on all the data in the DIR, we published two very impactful publications. One was looking at top 10 CT protocols in the nation for adult patients and then top 10 for pediatric patients.

[00:07:26] And I think giving that dose estimate or benchmark. And remember, it’s not really a dose. I should be very careful here. It’s not really dose to a patient because you’re looking at CTDI, which is more like a dose to a phantom. But you’re also looking at SSD, which is size specific dose index, which could be looked at to be one step closer to being dose to patients.

[00:07:46] And so for hospitals that have never done dose monitoring, this was one way of getting involved and checking how they’re doing. And then adjusting their protocols up or down depending whether they were, you know, lower in dose compared to their peers [00: 08:00] or higher in dose compared to their peers. So I would say that the DIR has had a significant impact on the nation in terms of dose monitoring and establishing benchmarks.

[00:08:10] Chris St John: Yeah, and so my understanding, and correct me if I’m wrong, is like the DIR aggregates data from, you said it’s just CT, right? It’s just CT. But from all these different healthcare facilities, were there any challenges about the standardization of that data across different institutions?

[00:08:29] Kalpana Kanal: Absolutely. So the way this works is when you enroll in the DIR, you have to go through the process of what’s called mapping.

[00:08:36] So, what is the HET CT protocol at University of Washington, which is my institute, versus, you know, the HET CT protocol at MGH on the East Coast, you know? How do we make sure that we are talking about the same protocol, right? So, you have to map it to some common terminology within the DIR to make sure you’re comparing apples to apples.

[00:08:56] So, that is a very challenging process, depending on how many [00: 09:00] protocols you have. And I know DIR is continually working on making that process easy. So I think that initially for us was a challenge trying to just map all our protocols. So if you have that right, right, then you’re comparing apples to apples.

[00:09:14] But then the other thing the DIR can do is the only information you can get is whatever is on the DICOM header. So the some of the challenges of questions I got when I was a chair was, you know, why are we not looking at at doses as a function of scanners that have modern dose reducing software like iterative using reiterated reconstruction or a reconstruction or more modern scanners compared to old scanners?

[00:09:40] Well, if it’s not available in the DICOM header, the Dose Index Registry cannot really tease out that data. That was one of, I think, the most common complaints I got when I was a chair of that program. So, I think there are challenges in doing this. And even when we did our papers, you know, there’s challenges in curating the data.

[00:09:59] You [00: 10:00] want to make sure your data is clean. There is no contamination with, you know, combo exams or, or the mapping not being done correctly. I think the one thing the DIR is relying on facilities is to do the mapping. You know, they are not looking at what mapping you’re doing. You’re doing the mapping and They are not looking at our image quality.

[00:10:19] So you’re sending the dose, but they have no idea what your image quality is because we are not sending them any images. So I assume your next question will be, what is one of the drawbacks? And I kind of already talking about that.

[00:10:31] Chris St John: You can get ahead of me, go for it. I love that.

[00:10:34] Kalpana Kanal: Right? So since we are already talking about this, it’s great to have the doses, but I don’t know that that routine head CT at the University of Washington versus the routine head CT I picked on MGH.

[00:10:46] Is what is our image quality? Even though our doses might be comparable, you know, is it what is my image quality? What tools am I using? What scanners am I using? So that definitely would be the other side of the coin. You cannot look at dose without looking at image [00: 11:00] quality, and that’s something the D. I.

[00:11:01] R. Didn’t have at that time, right? And I still don’t think we send any images and look at image quality. We only look at dose. But I think that is something that would make sense if you could look at both sides of the coin and then look at dose and compare it to the image quality. So I think there are some drawbacks, but I still would encourage anyone who does not have a dose monitoring software at their institution, this is a really easy way to participate and get involved and monitor your doses.

[00:11:28] Chris St John: Yeah, absolutely. And like, you know, regarding image quality, right? With the new CMS clinical quality measure coming up, I think outpatient is going to be required in 2027. So at some point, everybody’s going to start looking at their image quality versus their dose quite a bit, I feel like.

[00:11:45] Kalpana Kanal: So yeah, there is that CMS requirement.

[00:11:47] And I know that the dose index registry is probably going to figure out how to incorporate that or come up with a solution with their current platform and then sites who have dose management software. Are also, I think, the [00: 12:00] vendors have already started looking at trying to implement that in their software.

[00:12:04] So, yeah, it’s coming. We’ll see what happens.

[00:12:07] Chris St John: Yeah, we absolutely will. So, can, can we talk a little bit about those two, uh, research papers? I believe you were focused on establishing diagnostic reference levels or DRLs. and achievable doses for both adult and pediatric CT using DIR data. How did the DIR contribute to ensuring dose benchmarks were accurate and, you know, clinically useful?

[00:12:33] Kalpana Kanal: Dr. Raghuram G. Rajan Yeah, I see what you’re asking. So, you know, as I, again, I, I’m totally blanking on when DIR started. I think it was around I said 2004, 2005, but I think it might have been a little later. So when we had enough data in the database, right, then we said, okay, now it’s time to start thinking about establishing some benchmarks from all this data we have across the country.

[00:12:55] So for the adult population, we looked at about 101. 3 million [00: 13:00] exams, I think it was, if I’m not mistaken. And for the Peds, I think it was about 1. 5. I definitely should have looked at the papers before doing this podcast. I don’t remember. For a little while. But the adult, so we, what we started doing is, okay, let’s start with the adult, right?

[00:13:13] So we have all this data that’s sitting in our database. Let’s establish a benchmarks. What is a typical dose that the facilities, like minded facilities are seeing across the country? And what are our top 10 exams? So we started working with the DIR. I think the data curation took us probably a year. I was working with my other physicist who at that time was still working for the American College of Radiology, Miss Priscilla Butler.

[00:13:38] And so we started working with the DIR and say, okay, this is how we want to frame the paper, you know, send us some data. And so the data curation went on. Is this clean data? Do we see anything else? Do we need to drop some of these combo type of studies and things like that? That took us about a year and then we managed to get a really good statistics on the data too, right?[00: 14:00]

[00:14:00] No data is good without good statistics. And Dr. Ihsan Sehmeh was involved, Dr. Frush was involved in the pediatric paper, some of very well known radiologists and physicists. And after that data curation period, you know, and the analysis, we had a lot of help from the ACR statisticians and staff, Ms. Butler and I started writing the paper with the help of our colleagues.

[00:14:20] And we said, okay, after that paper was established, the first, the adult paper. Now you had a resource that was built, that established benchmarks built on a really large number of data. The more data you have, the less is your standard deviation. So the more solid, right, the data is. And so that benchmark was the first time in the United States that we established benchmarks for the top 10 exams in the country.

[00:14:46] And I thought that was great because now any site You can look at their CT head or CT abdomen or, you know, chest exam and compare it with these benchmarks and see how they’re doing. That’s a [00: 15:00] great starting point. I mean, it’s a great way to start your dose management. And diagnostic reference levels are basically established at the 75th percentile of your data.

[00:15:09] And if you exceed that, then you need to look at why is your data high. It really means that most of the facilities are. can get data less than that 75th percentile, right? So I think it was a monumental undertaking, but I think the result was so impactful on the population, not only in the U. S., but outside the U.

[00:15:30] S. Anywhere where people, physicists, do dose management. Having this resource was extremely useful. So once we got done with the adult paper, then everyone said, Hey, what about it’s because beads are in a more sensitive population. So we took that on after the adult paper and started working on that. But in beads, we also did one step further.

[00:15:50] We went, we did age and size because some hospitals Look at their protocols as a function of age and some do with size. So we produce, put both of them in the paper to [00: 16:00] give, you know, ammunition or give people a resource, if you will, to compare their protocols. That took even longer, like the data, because not a lot of people do PEDS and we had to definitely get a lot more data to get good statistics.

[00:16:13] You know, we had some guidelines that we followed when we were taking data for that paper and those all established or. Mentioned in the paper as well. So I think those were very impactful publications. And you know, even though I’m first author, I cannot take the credit for it. I had a whole team of people, right?

[00:16:29] The DIR people, other physicists, radiologists. So it really was a village that built this great resource out there. So I think it’s time again now for DIR to reestablish DRLs because it’s already been almost eight years for the adult paper and four or five years for the pediatric paper.

[00:16:49] Chris St John: Yeah, it’s so funny to hear, from my perspective as someone who didn’t even take high school physics, it’s so funny to me to hear like, PEDS was only age specific until [00:17:00] y’all decided to introduce size as well.

[00:17:02] I’m a 33 year old man and I know 15 year olds who are larger than me, like.

[00:17:07] Kalpana Kanal: Right, right, exactly. And as I said, some hospitals. Look at their pediatric protocol as a function of age. And some look at size, as you said, a 15-year-old might be a 300 pound patient, possibly. Right? Or a tiny patient. And so you have to adjust your protocols according to the size of the patient, whether it’s PS or adult.

[00:17:26] And I think these are just a starting guidelines, right? Okay. This is what we got from the 1.3 or 1.5 million data points. This is what we are seeing as the DRL, you know, the 75th percentile. In these papers, we also had the achievable dose, which is the 50th percentile, and most people should be in the 50th percentile range, but you’re saying, okay, here’s your range, 50th percentile to the DRL, 75th percentile.

[00:17:50] If you’re here, You’re good. If you’re over 75th or closer to the higher range, you know, you may want to look at your protocol. Why are you higher compared to the rest of the [00: 18:00] facilities who can achieve this target? So I think they were very impactful papers, but I know DIR will probably, ACR will probably now look at writing newer updated papers with more data coming in because you have to keep looking at this.

[00:18:14] Benchmarks over every three to five years to make sure they remain the same. It’s a moving target with new machines coming out, new bells and whistles for reducing dose, AI coming tools now for reconstruction. I think it’s time again that they will be looking at that again. I think you’re going to be hopefully talking to Dr.

[00:18:33] McKinney, who’s the current DIR, who can tell you what they’re thinking about for future.

[00:18:37] Chris St John: I’m curious, what are some of the biggest misconceptions when it comes to DRLs and how can facilities use them more effectively?

[00:18:46] Kalpana Kanal: What are some of the misconceptions? I think maybe people have understand. I don’t even know if there are misconceptions, what you’re saying.

[00:18:53] I think people do need to understand how to calculate them. I think that’s important. [00: 19:00] The methodology is important, and there might be some misunderstanding about the methodology. And people should be following, and I believe it’s ICRP 135, I think, which defines the methodology. Again, I don’t recall all the details, but So that might be an issue, what DRLs and how they are defined, and just for people to have tools and resources to help them figure it out, right?

[00:19:21] So if I’m not participating in the Dose Index Registry and I do not have these fancy dose monitoring software, how do I come up with DRLs for my institution, right? So we try to do that a little bit in the paper and in all the presentations I’ve given since then. You know, where do you start and how do you establish that?

[00:19:39] So I think the methodology is important for sites to understand. You could do this with a small number of patients, take your most common exam and do it. So I think in terms of misconception, I think understanding the process is important.

[00:19:51] Chris St John: So can you talk about where some of these smaller facilities should start?

[00:19:55] Kalpana Kanal: Oh, absolutely. I think a smaller facility that does not have a lot of [00:20:00] resources and don’t have dose monitoring software. I think the ACR, they could participate in the ACR, but if that’s still financially not feasible, I think they can start with their most common exam, right? And monitor 20, 30, 40, 50 patients.

[00:20:15] Look at the CTDI, the DLP, the SSDE, and then do the math, follow the Process, which I’m not going to go into detail here, but there is resources out there, right? So, and that’s where they could start, right? With one, one protocol, their most common protocol. Start the work to you. It’s manually totally possible in an Excel spreadsheet and see, okay, I did my head CT.

[00:20:38] Now, let’s see what my DRL and AD is. And then I can compare it to these papers, right, if they’re looking at beats or whether they’re looking at adult. There, you got one done. You have an idea. You have a handle on your doses, right? Better than not knowing anything, how your doses are doing. So that’s what I would advise small facilities to do.

[00:20:56] Start small. Take the most common protocol you have and the most [00: 21:00] impactful that impacts most of your population and start doing this. And once you do that for one, it’s easy to extend your range, pick your top ten protocols if you like, you don’t have to do this for every protocol. And then expand your range, you do it for one facility, then you do it for your second facility, third, how many ever you have.

[00:21:18] Because then, not only is it important that your doses at one hospital are optimized, but that the doses at your other hospitals, if you’re managing more than one hospital, are also optimized. So I don’t think it’s difficult to do, because there are many resources out there. ACR itself, the DIR program, is such a resource, but there are a lot of papers, there is presentations, NAAPM, and so on, where people can find resources.

[00:21:43] You can talk to your physicist colleagues all over the world. I’m working with one right now in Nigeria to establish DRLs for their doses. Not difficult to do, you know, if you know there are resources out there.

[00:21:56] Chris St John: Absolutely. I think that’s exactly what I was looking to hear. And yeah, I mean, it’s [00:22:00] funny, the more and more I do this podcast, the more so many answers feel just close to just Like, everyday life things, just within the veil of medical physics, right?

[00:22:11] That first, you know, the journey of getting better at something, just that first tangible step, right? I, there’s a, there’s a silly kid’s cartoon that I love, where somebody says, Sucking at something is the first step to being amazing at something. And I’m not saying people are terrible with their doses, but just like, Taking that step is like, So, so necessary.

[00:22:32] I’m trying to learn how to sew. My first thing was like, I want to run a seam. One seam. Okay, I’m done. I did it.

[00:22:38] Kalpana Kanal: Yeah, and you know, this whole thing started many years ago where all the spotlight came on CT dose because of some high dose cases that were published in newspapers and that put the whole focus, oh my god, that we are You know, we’re giving cancer to our patients, we’re doing this, we’re doing that.

[00:22:54] Well, I wouldn’t quite agree with that. I think, yeah, maybe our doses were a little all over the place. [00: 23:00] But I think we’ve come a long way. I think we are almost guilty now of going the other direction, maybe that we are too low. I think the word to use here is optimized, right? We want to optimize our dose.

[00:23:11] We don’t want to go low. It’s, I don’t want to go low. I want to do what’s best for my patient. That means getting the best image quality I can get at the minimal dose possible. But that doesn’t mean that I’m competing with anyone to be the low dose. I want to be an optimized facility, right? And that’s what accreditation helps us with ACR, right?

[00:23:29] They look at our doses, things like that to say, okay, are your protocols okay? But I’m really worried that we might be going the other direction, you know, how low is low, how much money do you want to spend trying to keep going low? So I think the key here is finding balance. Start your first step, you know, get a handle on your doses, optimize them.

[00:23:47] And one thing I want to say is you may not be able to optimize it for every patient or every protocol because it depends on the need of the radiologist, right? And it depends on what vendor CT machine you has. Is it the old [00: 24:00] one? Is it the new one? So I think doing the best with keeping the patient in mind but at the same time what the radiologist needs, extremely important.

[00:24:08] We are physicists. We don’t review images. The radiologists do. We don’t want them missing anything. So I think optimizing the dose and not trying to say, No, you cannot read at this. We’re going to lower. No. You have to work with your radiologist. You have to work with your technologist and get that balance where you are confident that your doses are optimized and your radiologist is confident.

[00:24:29] In their diagnosis when they’re looking at those images. That is critical.

[00:24:32] Chris St John: Yeah, for sure. I’ve been just surrounded by these conversations more and more these days. I’m like, if I ever get into the CT, crank up the volume. You know what I mean? Like, give me that HD 1080p image quality. Zap me. I mean, I want to see everything.

[00:24:45] Kalpana Kanal: We don’t want to do that either. I know.

[00:24:50] Chris St John: I’m teasing. Okay, so after playing such an integral role in DIR’s early days, you have since stepped back from the initiative. What [00:25:00] kind of led to that decision and where are you focused now on your research and your professional work?

[00:25:04] Kalpana Kanal: Yeah. So I go to I was very fortunate to become a trustee, a Diagnostic Medical Physics trustee for the American Board of Radiology.

[00:25:12] And I had to step back from the ACR because it was really difficult to manage both those leadership positions. The ABR trustee role is pretty demanding of time, but it’s a very fulfilling role as well. So I’ve been doing that since 2017, and that’s why I stepped back from DIR. I mean, I’m still involved with ACR.

[00:25:30] We’re still trying to write some. papers. I’m very much talked to all the leadership there now as well, but I’m not actively involved right now. My focus has been in terms of work has been on a very interesting new path that not a lot of people in physics are, which is forensic CT. So what that is, is basically using CT.

[00:25:51] So you’re leveraging technology To determine cause of death or trauma or whatever is going on in a cadaver. So instead of, imagine here, [00: 26:00] instead of doing an autopsy, if you could have that cadaver or that dead person go through a CT and you can figure out what is the cause of death. and don’t have to cut open the body, that is significant.

[00:26:12] So, I’ve been working on this for the last six, seven years. So, my chairman, Dr. Dushan Sahani, when he came to University of Washington and he was talking about some of his far thinking ideas and his vision, this whole world of forensic CT totally resonated with me. And I just went with it and he gave me his blessing to move forward with it over here at University of Washington.

[00:26:34] So we started working on this and we have to today, we have scanned about more than 150 cadavers on our Siemens CT scanner. You can do it on any CT. It’s just that our semen CT scanner is right one floor over our medical examiner’s office. And we have scanned 150, 160 of them. And about 45 to 50 percent of the cases, an autopsy was not needed because the CT gave them the [00: 27:00] answer to the reason for the cause of death or the trauma or whatever the reason might be.

[00:27:04] And these are cadavers that come, you know, decedents as well. It’s called, uh, from suicide, trauma. Accident, you know, natural death, stuff like that. And that has been a fascinating, fascinating pathway that I have taken. At AAPM, I presented because I want people to be aware of this. In America, here in the U.

[00:27:27] S., not a lot of sites do this. But outside America, a lot of sites do it. Japan does it. Australia does it. There’s England that does it. And why not here? I mean, just imagine the impact on community if you do not have to cut up a cadaver, right? Think about the cultural and emotional sentiments that people have and saying, you know, my relative suffered enough and now you want to cut them open, right?

[00:27:48] I wouldn’t want that for my relative. So I think it’s going to have a far reaching effect. We are publishing on it right now. It’s in the process of being published. I’ve given several talks at physicist meetings to bring [00: 28:00] awareness. And like, the main common question I get is, why not, right? Why aren’t we doing it more in the U.

[00:28:05] S.? And I think it is an issue of billing and things like that for anything, because radiologists have to read these images. We work with pathologists on doing this. So we have established this program here, and we did such a great job of convincing our medical examiner colleagues that we help them buy a CT scanner, which is going to be installed in the medical examiner’s office in a couple of months.

[00:28:26] And then they’re going to have every cadaver go through it. For my project, we did selective, right, because we didn’t have the bandwidth. But now imagine, the whole new world opens up. The other reason this is great using technology, but from a ME perspective, there’s a shortage of forensic pathologists. So there aren’t a lot of people available to do autopsies.

[00:28:47] They’re backed up for days. Using technology to help. Why not? Right? So this is something I have been very interested in for the last four, five years, and that’s what I’ve been working on to establish a forensic program here at [00: 29:00] University of Washington. Very, very interesting to me.

[00:29:03] Chris St John: Can I just ask? I’m so curious.

[00:29:05] So, I mean, obviously, there’s different cadavers coming in, and maybe you have clues of, like, when it comes to selecting protocols for these cadavers, is it always full body scans? Is it case by case? Like, I’m kind of just curious about how you approach this.

[00:29:22] Kalpana Kanal: Correct. So, we had the same questions when we got started.

[00:29:26] And I think we have now worked so much in it that it’s getting more fine tuned that the forensic pathologists who order these CTs from the medical examiner’s office actually give us guidelines. They say, you know, we’re interested in the pelvic area, so can you just run a pelvic CT? Or we’re interested in chest, abdomen, pelvis, or we’re interested in the head.

[00:29:44] So now we are actually moving away from doing whole body, and that’s actually good because the scanner, if you’re doing whole body. You’re heating up the scanner, you know, you’re creating tons of images, which no one is necessarily going to look at. You’re bogging down your

[00:29:58] Chris St John: Right. How, like, how many slices do [00:30:00] you have to go through for a full body scan?

[00:30:02] Kalpana Kanal: Exactly. A lot. You know, it also depends what slice thickness you’re reading at. So we have now actually worked so well with our program and our medical examiner’s office that now they give us guidelines. They say, okay, for this cadaver, you know, say it was a car accident. So it’s an MVA. We suspect there’s trauma in this area, can you do a CT here, or we suspect this is going on.

[00:30:22] There’s a lot of conversation going on between the forensic pathologists and our radiologists. They get some guidance, and so we are now more focused in what we do. And we also actually do conferences every other month, I think it is, or every month, where, kind of like an M& M conference, where we are looking at these cases, and we are talking, it’s so interesting, because you’re talking The radiologist talking, hey, here is a finding on the CT.

[00:30:47] And then the forensic pathologist is putting their gory images with the blood and gore and anatomy, which now you’re used to seeing and you’re trying to correlate the two, right? which is so fantastic. So they learn [00: 31:00] a lot from our radiologists, and our radiologists learn a lot from the forensic pathologists.

[00:31:04] What are their needs? What are they looking for? And I would say we are a good example of the success for someone who’s trying to establish the program. As I said, there are other programs. One in Albuquerque, OMI at University of New Mexico is a very well established program for years. They scan every cadaver that comes over there.

[00:31:22] So, you know, we are learning. From other people, and I think we have been very successful in doing this, and I hope to, you know, see more happening in the future. And then you can apply this to education, right, medical student education. You don’t need to cut up bodies to look at anatomy, you can look at images.

[00:31:38] CT images and see what it looks like and things like that. So there’s a limitless potential for taking this to any other area you want to.

[00:31:48] Chris St John: And so I can’t help but wonder with you and your background in dose optimization now working on forensic CT. So, you know, your past [00:32:00] experience with D. I. R. now working in Forensic CT, like, did you have any observations about dose or image quality that you kind of picked up while working with cadavers?

[00:32:10] Kalpana Kanal: The one thing is that dead people images look very different than live people, right? There’s no blood flow, and the more further away from death you’re imaging, it’s going to look different. So I know that the dose is not a concern, they’re dead, right? Yeah, so you’re not worried about optimizing dose. But at the same time, you’re not overexposing them either.

[00:32:28] We are really focused more on the radiologists and what image quality they want, right? So we have really tweaked our Protocols because we use automatic tube current modulation and the radiologists have been happy with it. They haven’t told us to tweak anything because again, it’s hard for them anyway to look and relearn what anatomy looks like in a dead person because there’s no blood flow and things like that.

[00:32:52] And that’s what they’re focused on. They’re not focused on dose or anything. So that’s not even come up. But we might do some projects with our residents in the near [00: 33:00] future to look at, you know, all the 150 patients we have done so far. What has been the dose? Curious, right? If there was a 300 pounder or a 100 pound patient, what is our range of doses for these cadavers?

[00:33:12] Is our radiologists who are reading these are so far seem to be okay with image quality. They haven’t told us to increase the dose. So it’s a different question. The focus is more on, you know, are we finding what we need to find with the physiological changes in cadavers because they are not alive more than worrying about dose.

[00:33:31] So the focus is a little different.

[00:33:33] Chris St John: Yeah, I wasn’t thinking that you were focused on dose, but I was just curious about Yeah, like what your approach was when it came to image quality.

[00:33:42] Kalpana Kanal: Yeah, we took the same clinical protocol and we’re using this we probably tweaked up our dose a little bit because we we were not worried about harming anyone.

[00:33:51] So maybe 10, 20 percent, which is not significant. And so we really haven’t deep dived, if you will, into the dose aspect of [00: 34:00] these 150 or so we have done so far. But that is something we’re looking, starting to look at now that our program is getting a little bit more established. We can now start looking at research projects along all these data that we have.

[00:34:11] Chris St John: That is way more interesting. I mean, it’s not that it wouldn’t be interesting, but you had me just like sucked in for a minute.

[00:34:18] Kalpana Kanal: Yeah, it’s been the response I’ve received at any audience. I’ve, I’ve presented this to physicists. I’ve presented this to technologists. We have, we, our paper is getting published in Jcat.

[00:34:29] It should be coming out soon. I have had that response. This is so interesting. Why don’t we do it? Well, yeah, why don’t we do it right in the US? So it’s a question of billing and time and the radiologists getting compensated to read these images.

[00:34:43] Chris St John: Yeah, absolutely. Wouldn’t, I mean, someone performing an autopsy, like, there’s billing and stuff involved in all of that as well.

[00:34:49] But yeah, no, it’s fascinating to hear and it just makes so much sense. Totally.

[00:34:54] Kalpana Kanal: That’s what I say.

[00:34:55] Chris St John: That is beyond fascinating. Dr. Kunal, thank you [00:35:00] so, so much for joining us today on Frame by Frame. It’s been an absolute pleasure having you.

[00:35:04] Kalpana Kanal: Thank you so much, Chris. That was a lot of fun. I didn’t even realize that we have a formal podcast going on here.

[00:35:09] It was like a conversation, so I really enjoyed it. Thank you very much.

[00:35:13] Chris St John: Good. Thank you so much. Frame by Frame, rethink Imaging is brought to you by eLog. Here you’ll find engaging interviews with thought leaders, experts, and patients sharing stories that showcase the transformative power of medical imaging.

[00:35:29] To discover how eLog is rethinking imaging and healthcare, visit elog.com. Be sure to subscribe to Frame by Frame, rethink Imaging on Apple Podcasts, Spotify, or wherever you listen. And from all of us here at MOGs, thanks for tuning in.

Questions from This Episode

What is the ACR CT Dose Index Registry?
It is a national ACR program that collects de-identified CT dose data from participating facilities, from small community hospitals to large academic centers. Sites map their protocols to common terminology, submit dose indices such as CTDI and SSDE, and compare their numbers against peer facilities across the country. Dr. Kanal describes it as the most affordable way for a facility without dose management software to start monitoring doses.
What are diagnostic reference levels and how are they set?
Diagnostic reference levels are dose benchmarks set at the 75th percentile of collected data for a given exam, with achievable dose at the 50th percentile. If a facility exceeds the DRL, that is a signal to investigate the protocol. Kanal’s team used registry data, about 1.3 million adult exams and around 1.5 million pediatric data points, to publish the first national DRLs for the top 10 CT exams in the United States.
Can CT really replace an autopsy?
Often, yes. In the forensic CT program Dr. Kanal built at the University of Washington, more than 150 cadavers have been scanned, and in 45 to 50 percent of cases the CT identified the cause of death without cutting open the body. Countries including Japan, Australia, and England already use postmortem CT routinely; billing and radiologist compensation remain the main barriers in the US.