Sam and Meg Lubner on Careers in Cancer Care, How Oncologists and Radiologists Can Best Communicate, and How AI is Helping in Cancer Diagnosis and Treatment

Meet Sam and oncologist. The general sense of altruism and the feeling of doing something really cool and really cutting edge every day was the kind of thing that made medicine kind of a no-brainer for me. And meet Meg, a radiologist. I joke with the kids that it gives me this sort of assisted superpower of extra vision so I can see inside people and it's awesome. You know like all the different imaging what altruism we have. Megan Sam are married and they're the guests on the science for a podcast today. Hi, welcome to the Science Fair podcast. I'm your host Susan Keatley. I'm a PhD chemist, writer and I love talking to scientists. On the Science Fair podcast, I aim to bring you conversations with scientists doing fascinating cutting edge work on all kinds of interesting phenomena ranging from physics to chemistry to biology and even the nature of science itself. In this third season of the podcast every other week two episodes will come out. On Mondays there will be a shorter 10-minute episode linking the scientists research to what's happening in the classroom and then on Thursday the full length interview. So come along and tune in for some science fair. Our guests today are Sam and Meg Lugner both are joining from Madison, Wisconsin. Sam is a hematologist and oncologist at the University of Wisconsin Health and an associate professor at the University of Wisconsin School of Medicine and Public Health where he directs the hematology and medical oncology fellowship program. Sam specializes in gastrointestinal malignancies. Meg is a professor of radiology at the University of Wisconsin School of Medicine and Public Health in the section of abdominal imaging and intervention. Meg works in the field of radiomics, a field focused on the extraction of quantitative information from diagnostic images and her research interests include new technology in CT scans which means using radiation like x-rays for instance to create detailed cross-sectional images of the body. Sam and Meg are also friends of mine from decades ago. Sam and I were in the same class as undergraduates at Princeton and we both found ourselves in Madison, Wisconsin, post-college. I was a graduate student in the anthropology department and Sam was in the medical school where he met Meg and Meg had this incredible fellowship called the NAP House Fellowship where she was living in a house with something like 12 other graduate and professional school students and later I got to take part in this fellowship with Meg's expert guidance on the application process and here we are today recording this podcast. Sam and Meg welcome to the show. Thank you. That's so happy to be here. I'm so grateful for the chance and about the NAP House I feel like for the listeners imagine like the old real world just all with PhDs you know okay so very much like that. Yes oh my gosh now like particular memories are filtering in. So Sam and Meg can you start by telling your listeners how each of you decided to pursue careers in medicine and what led you to the roles you have today. I was a history major in college and was really thinking about being a sports writer and did sports radio and thought that was really cool and kind of kicked the tires on that job and realized oh my gosh that is long hours and really kind of thankless for very little pay and there are very few jobs and I looked at what was upsetting about that and also it just sort of felt like it didn't help anybody. So then I was like well I'm okay with the long hours and hard work and being sweaty and not sleeping but I wasn't okay with the very few jobs and the low pay and things like that and thought hey what what's similar and what would allow me to help people and hey med school sounds good. I'm sort of kidding on some of that but largely the general sense of altruism and the feeling of doing something really cool and really cutting edge every day was the kind of thing that made medicine kind of a no-brainer for me. As I reflect on it I had no idea what a career in medicine entailed in terms of length of training and what a job actually looked like but as I get further into it and realize that I do have a lot of control over the stuff that I do and the work that I pursue it has ended up being a real great career for me. As I started in my career I was really heavily focused on GI Cancer Research but I really loved teaching and I really loved mentoring students and residents and fellows and as I went further in my career I realized that I wasn't I wasn't as good at research as I thought and I didn't like it as much as I thought and as my career progressed I got into student advising and student services and I'm now the assistant dean for students here at UW and have found that to be a really enriching part of my career where I get to not only help my patients with their clinical presentations but I get to help students navigate this pathway of education and mentorship that brought me to this place. I guess for me it's hard to remember exactly when I decided that this was what I was going to do I feel like I always loved science which Sam probably did too and in the spirit of this podcast I had some amazing science teachers in middle school and high school who really cultivated that interest and I also remember loving physics which I consider kind of a combination of science and math so definitely I had an amazing high school physics teacher who I still keep in touch with and so you know I liked all those things and you know it was a long time ago when I graduated from high school and back then they had some programs in Wisconsin where you could apply and could be conditionally accepted to medical school out of high school if you completed you know certain requirements and maintained a certain grade point average and the idea behind it was that when they were able you maybe to take a less traditional majoring college as long as you did the prerequisite which Sam kind of did anyway so a lot of us I was still science majoring college but some of the other students who went this way did things like a Spanish major or a history major or so it made them in some ways a more well-rounded person approaching medical school I think but I so I had applied to that but accepted and so you know even when I finished high school I was thinking that you know medicine was just a great way to pursue a career in science for the reasons that Sam talked about I feel like the mission is very built in for me it feels like a very practical application of science you know great science that's being done you know we can translate it to patients and improving their outcomes you know which is very satisfying application of science for me and a very tangible application of science for me and so that was kind of so I think I had decided on medicine fairly early on you know went through college as a science major loved my science classes loved physics loved physical chemistry and took a year off before medical school because I had decided so long ago that I just wanted to make sure that this is actually what I wanted to do and worked as a receptionist and a clinic really honed my phone skills which is very useful in medicine and then ended up going medical school after that and you know during medical school and during residency I found that I loved everything you know like every part of medicine I loved every rotation I did I'm like oh I like this I think I'm gonna do this and so it just happened that radiology was probably the last rotation I did before I applied and I had some great again mentorship in radiology where they said this could be a great career for you if you like physics if you like physical chemistry you know you should think about about radiology and you know it was not what I ever would have pictured for myself I I love patients I love taking care of patients and for those that know sort of the you know traditional idea of a radiologist is someone who sits in a dark room and doesn't talk to anyone and certainly there are some days when I do that but there's a lot of days where we're doing image guided biopsy or image guided treatments of tumors so there's so it's changed a lot evolved in a lot of ways and I love the abdominal imaging part of my job I love the intervention part of my job it lets me do kind of both things and although I don't see patients in the longitudinal way that Sam does I still get to see them you know sometimes they may come in with pain you know we read their imaging study we say there could be some sort of problem here maybe we should sample it and then I do the biopsy and then I see the results so it's an incredibly educational process for me to see what what it looked like when it started to sample it and find out what it actually was. And I feel like it's very useful for the patient because then we can sort of triage them to the care that they need. And so I feel like it's been a very satisfying role. It's not quite, I'm not quite seeing patients in a longitudinal way that's amaz, but it's very, I feel like we really have become an integral part of the team. And I joke with the kids that it gives me this sort of assisted superpower of x-ray vision. So I can see inside people and it's it's awesome, you know, like all the different imaging what-alities we have. And now, as you mentioned in your intro, you know, we're doing so much imaging for a given purpose, but there's so much data there that maybe we're not using. And so that's something we're really looking at now. Are there opportunities to screen for other types of diseases with this data that we're getting for some other reason? And that's been a really exciting part of my job too. I love the idea of you having x-ray vision. Yeah, I always joke with the kids. Or I'll say to Sam, he'll talk to me about a patient who'd be like, "I don't really know what's wrong with his patient." And I'll say, "Well, what we need to do is put them in the truth machine, which is the CT scanner." And that will know. And although we may not exactly pin down what it is, we can at least say, "Well, here's what I think's involved. Here's where I think the problem is, let's look there." You know, so I, yeah, I do love that. It's just such a neat idea, you know, to run the person through the machine and then we can kind of look inside without, you know, in a very non-invasive way. Right. And I wanted to ask, you mentioned you had some wonderful teachers, including a high school physics teacher that you're still in touch with. What made that teacher or some of the other ones so great? Well, I think one is, you know, he had a very natural enthusiasm for science. And I feel like when someone is really genuinely enthusiastic about something, it's contagious, you're infectious, you know, you're kind of like in a good way, not in a, you know, COVID kind of way, but like, you know, and so I just feel like he was just so excited about it. I'm like, "Wow, this must be really cool." And you know, he used very practical examples that we as students could grasp. But the other thing that I loved about him is I just felt like he was really on our wavelength, meaning he was very interested in us in people, whether we were interested in physics or not, you know, like he just was very, one of those minds that's just sort of genuinely curious, engaged with gathering knowledge of all kinds, whether it's scientific knowledge or, you know, knowledge about you as a person, you know, just really bubbling over with enthusiasm all the time about everything, you know, and I just felt like it was a really magnetic, you just really couldn't help but be brought into it. And I just thought he did such a great job of showing how elegant physics can be. I feel like it's kind of an abstract concept when you're a high school student, like, what really is physics, you know, like, and it would, just the way that he presented it was just so beautiful and elegant. It was just hard to not love it. I thought maybe not everyone in the class felt that way, but, you know, and I just feel like he still, you know, he still plays pickleball with my parents and will ask about me and is really excited, genuinely interested and excited, you know, like many, many, many years later. So I just feel like his interest in science and everything is just, he's just, I, it reminded him, Ted Lasseau a little bit about being curious and not judgmental, like that is his approach to life. And I just, I just was totally taken into my, I, I thought he was amazing. That is lovely. So I think each of you alluded to some aspects of your days already, but I would love if you could each walk us through what a typical day in your career, in your job looks like. I think your days are easier to walk through than mine. I'll let you go first. Well, in abdominal imaging, you know, we have a lot of different imaging modalities that we can use. So we can use CT, computer tomography. And as you mentioned, we use ionizing radiation for that to generate images. And again, it's a really elegant mechanism the way it works. It's pretty amazing to me that it works, but that's probably our workhorse imaging modality. So a whole spectrum of different clinical scenarios may come in and that may be where we start. But we also do ultrasound where we generate the images with sound beams again, an amazing application of sound waves to generate these images. They look very different. When I first saw an ultrasound, I had no idea what I was looking at, but I feel like they're very beautiful once you kind of know what you're looking at. And ultrasound is a great modality for doing image guided interventions. So if we were going to biopsy a sample, a tumor, and we might do it under ultrasound, we also have MRI where we generate the images with magnetic waves. Again, a major physics breakthrough and just amazing the kind of tissue characterization you can do with MR. And then we also do fluoros. So really that is just x-ray, but it's a very real time modality. So we'll often have patients drink some sort of contrast agent like barium that we can see on the x-rays. And then we can kind of watch it move through. So if you have something like the esophagus or the stomach or the small ball and you want to see how do the muscles work together, are things moving through in the way they should? Is there a blockage somewhere? You can turn on fluoro and have them drink and just watch that contrast agent go through. We can also place feeding tubes under fluoro, we can do other procedures under fluoro. So we have all these different modalities and then we also do CT colonography, which is a different thing. It's screening the colon and for cancer, but it's done with CT. And so that's another rotation that we do when we're on GI. So when I became an abdominal imager, I was afraid maybe I was giving up this whole screening, which is kind of a different idea and a diagnostic imaging, but we do a little bit of all of it. And so on a given day, I may be assigned to a service, so I might be assigned to CT. And I'm going to read all the CTs that come in during a given time period. And it could be cancer followups, it could be ER patients, it could be some of our sick in patients who maybe had a surgery and now have a fever. So all different kinds of stuff coming through transplant patients, etc. So that might be on CT for that whole day. And then the next day, I might be on ultrasound. And so on ultrasound, we would do a combination of diagnostic work where we're doing ultrasound for someone with right up or quadrant pain because maybe they have gallstones. We want to see if they're causing a problem, but we'll also do procedures. So we may sample tumors, we may remove fluid that's accumulated in a place that it shouldn't have or something like that. And so ultrasound is a really exciting service because you're just doing some diagnostic some procedural work. It's always touching patients. Yeah. And I get to see patients sometimes they're same patients, you know. And so sometimes on other days I see same patients too. Another day I might be on a mar. So I'm reading all them Mars that are coming. Another day I might be on GI, so I'm doing all the floral studies. So it's a really cool mix of stuff. And then maybe one day I'm not on clinical service, I'm academic. And so on that day, maybe I will, you know, in the summer, we do a lot of summer research with medical students. So maybe I'll meet with the students. Maybe I'll work on writing a paper about something that we're working on or maybe I'll meet with, you know, people at the hospital about CT operations. You know, so that kind of stuff. So there are some days where I'm not doing clinical work. And then, you know, and maybe, and and on every service we have learners as Sam said. So, you know, we call where we read, you know, where we read studies the reading room, which is usually kind of dark because we're looking at the images and you see better in the dark. So the kids laugh because a lot of times I go into a room and I just don't turn the lights on. It's just my natural state. And as I get all good, it's a good look for me. The darker probably the better. And so the reading room now we have this like giant combined reading room. And the reason we've done that is we want to have this sort of like community structure. So whether I'm on CT or ultrasound or MR or GI, we're all in the same room. So who's so so if I see a case that's kind of strange or I'm not sure what it is or it's just interesting, I can just turn over my children and be like, Hey, my colleague, look at this. What do you think about this? You know, we can talk about it. Excuse me, a medical student that's in the room. I would call them by their name. Not to say that. But, you know, let me show you this interesting case. You know, my PG my first year resident, let me show you this. My second year resident, I just reviewed your case. I think you should look at that again. You know, I think there's something there that you didn't see and then I want you to tell me what it is. You know, and then we've got, you know, referring providers coming in. They're like, my patient had an MR this morning. Can we go through it? Let me tell you a little bit more about them, which is like one of my favorite things, because I feel like the more information we have, the better read we can provide. And of course, I love our referring providers and when they come in and in the time of COVID that stopped for a while. So it's really nice when they come in again. So so the reading room is just like this bubbling, living place kind of where there's like all kinds of learners, all kinds of faculty, you know, lots of exchange of ideas. And it does get a little bit loud and a little bit chaotic sometimes, until you know, sometimes we have to tone it down a little. But you know, like for example, on the reading room the other day we were having this conversation where one of the trainees said to me, what is your most favorite and your least favorite organ in the abdomen, which is a very thought-provoking question. Very thoughtful. I've never known that in, you know, so we had a lot of discussion about what our most in least favorite is. You know, so you get all these. It's just a really cool exchange of ideas. I find it very energizing. But my day is different than Sam. So. Well, it just sounds like that guy was hitting on you. I'm not sure. No, it was a female resident. All right. All right. Well, just before Sam goes, I just want to jump in and thank you for that amazing description of the reading room. I had no idea. I actually, you know, based on not much, just pictured what you originally described, like someone looking at images in a dark room. And definitely there are some reading rooms like that. Arzus is just not like that. And recently, we had a young undergraduate. He's actually the son of one of my high school classmates, who's interested in reading room. Very forward-thinking. He's like a freshman or sophomore college. And I didn't even know what radiology was when I was that age. But he's like, you know, I think I'm interested. So I was like, well, why don't you come to the reading room and see what what it's like. And you can imagine this, you know, whatever it is, 19-year-old thrown into this mix of like discussion. There's just so much talking and back and forth and his dad texted me after. He's like, "I don't really know what happens in the reading room, but he came home very, very excited." You know? That's amazing. Anyway, it's a funny, funny thing. Yeah, it sounds really stimulating. And also, I mean, there are so many conversations about remote work versus office work, but I think what you're describing happening in the reading room, it really touches on one of the important parts of having people all together. Absolutely. You're having these exchanges and these conversations, and especially for younger people, learning, there's nothing that would replace, you know, that you look over your shoulder, "Hey, take a look at this," or, you know, even the, what is your favorite organ question? Like, I'm sure it just builds rapport. Totally. People are learning new things, so I really think that is a wonderful, visual that you've given us. And I will say, the community reading room is new. We've only had it for a year. We used to be kind of in different reading rooms, and that's exactly why. We just really wanted to have more kind of a spree decor. We wanted to be able to share ideas a little bit more freely. I just feel like, as you say, there's no real substitute for proximity, for a lot of these kind of exchange of ideas. And even though I've been practicing for almost 20 years, I'm always learning from my colleagues, you know? Like, even like, you know, they're training the other day, was like, let me show you this keystroke shortcut on packs that you don't know, old person, you know? And it's just, there's always a back and forth of teaching, which I have loved. And I still, I feel like it's good for our younger faculty, but it's also really good for me, because I feel like I'm always learning, seeing more cases for all of us is better. So anyway, sorry to talk so. I'm wondering if it's okay. Oh my gosh, that's great. No, I mean, so my day and megs days are, really kind of co-dependent on one another. I really depend on the facts and the data that Meg gives me in my clinical life. So I have, with my, the various hats that I wear. So I have my program director hat, which is 30% of my week. I have my dean of students hat, which is 40% of my week, and then my clinical hat, which is the other 30%. I also have, yeah, theoretically, I also have a bunch of other hats that bring me to about 130% but that's a whole lot of things. But my clinical life, I think, is the easiest to explain. And what that involves is I see a slate of patients who are either on treatment, under surveillance after treatment, or are newly diagnosed and we're making a plan. So the folks who are on treatment, those are people with cancers that have been diagnosed. And I'm either giving them chemotherapy pre- or post-operatively with the intent to reduce the risk of recurrence. Or I'm giving them palliative chemotherapy for metastatic disease, meaning I'm not curing them, but I'm buying them more time and trying to alleviate the symptoms of the cancer with the treatments that I prescribe. And there's a lot of treatment monitoring that has to happen because, you know, this is an aspirin, right? Like this is chemotherapy and little tweaks in doses mean a great deal to the patient's side effects over the next period of time. So I spend a lot of time getting to know somebody at an initial visit, trying to find out what are their goals, what are the things that give the joy, what's coming up in their lives, so that I can try to fit in the treatments that I'm proposing into their lives, right? And so that upfront sort of getting to know you in exchange of expectations is really important so that when they're on these on treatment visits, it can be relatively straightforward. How have the last couple weeks been? You know, specific side effects that I give patients with chemotherapy are things like nausea, vomiting, diarrhea, neuropathy, and immune suppression. So I'm always asking about those symptoms very specifically. And kind of in my head, checking the list of like, okay, we don't have to fix this, we don't have to fix this, we can just keep going in this way, or, ooh, all stop. Let's figure out what the cause of those symptoms for the last few weeks have been, and Taylor our next course of time between this visit and the next one to try to make that symptom better. And then the surveillance visits are folks who have been through all that stuff and were monitoring for disease recurrence. And both the folks who are on treatment and the folks who are monitoring for recurrence, I rely heavily on radiology because often either recurrence or progression is totally asymptomatic. And so I need a look inside the person at their anatomy and at their tumors and at specifically the size of their tumors to know are the treatments that I'm prescribing doing what I want them to do. And the new patient visits, the reason why I saved that for last is because it dovetails the most with sort of what Meg was talking about about a team-based environment. One of the things that we have set up here is we're co-located with our surgeons. Right, so if I have somebody with a cancer that at some point is going to need surgical resection, the surgeon and I meet the patient on the same day and say, hey, this is sort of what we propose. We would propose some chemotherapy followed by surgery or some surgery alone or surgery followed by chemotherapy, whatever that is. And just like Meg turns to her colleague and says, hey, you know, what's your favorite organ? I can say, hey, we're through four months out of six of planned chemotherapy. I need you to start looking for O.R. time for this guy because he's almost done. Or, hey, he's running into or she's got this refractory nausea and vomiting. I'm worried her intestines might be blocked up. Can you help alleviate that surgically? Do we need to change course? And honestly, I think we learned in because our previous model, we were not as closely located. But I would say both intra and post pandemic, we learned that there's no substitute for that face-to-face communication to say we're working together for the benefit of this person. How can we maximize my skills and yours to make sure that this person has the best outcome? That's so interesting. It's interesting that that came out of the pandemic. I'm assuming you. What did it look like before the pandemic? Yeah, largely it was phone-based. Like, hey, I saw a guy today, you know, call my surgeon up and say, I saw this person today, they're four months in. But, you know, in a busy clinic day, I don't have time to stop what I'm doing and wait for them to pick up or, you know, use the pager or send a raven or whatever to communicate to them where this person's at in treatment. So, the electronic health record has helped a lot in breaking that down. And certainly, when we didn't have the physical proximity, we would rely on those tools a lot. But I would say in the last few years, the actual being together has been a quality of life-improver for me, for my colleagues, and I think improve the kind of care we deliver for our patients. And I will say, just as the caveat, I think COVID did bring us some innovations that have been helpful that we still use. So, we do occasionally have a remote rotation, you know, so there is a little bit more reading from home. There's a little bit more flexibility than we used to have, but it's not every day. And I don't think it should be every day because. In video visits have changed right, you know, America's a big country, Wisconsin's a big state. So, I've got patients who are on oral chemotherapy who get their labs checked in Rhineland or Wisconsin four hours away. And I can do all of that follow-up remotely. Now, I can't listen to their heart and lungs, but I can see their white blood cell count and talk to them about stuff. And if there is something that feels like it needs an in-person visit, we can do that. So, yeah, the technology has changed the practice of medicine, a great deal. Some for the better, some for the worse. Yeah. The only other comment I would make that SAMCOT underplay is you can imagine that these patients sometimes are meeting SAMCOT and incredibly difficult time in their life when they may be receiving a really devastating diagnosis and their prognosis is variable. You know, maybe it's something that they're going to get through. Maybe it's not. And SAMC is great at taking the time to have these conversations with the patient, both, you know, to set realistic expectations, which can be really hard and feel hard for the patient, but also helping them understand that he's going to be there with them. And so one thing I always say about SAM and he's one of the best doctors that I know is that number one, he's a great communicator. And number two, one skill that he has is even if he has no tools to offer this patient, he will definitely care for them and they always feel that. You know, so I just think it's hard for him to describe that. in his day, but like these are the conversations that he's having and you know they're just these doctors are like very special people that are taking care of patients at this time in their life. I'm so glad you said that, Meg. And I'm not surprised at all to hear that about Sam. And I think it just kind of calls back to what you both opened with was that there's a mission in this work. Of course it's science-based, but there's something bigger. Yeah, it is so deeply personal and it is the kind of thing that the news that I have to deliver is often bad, is often the worst. At the same time, there is a certain amount of pride I take in being able to deliver that kind of news in a way that demonstrates that I really do and genuinely care for these folks, even if I can't fix the problem that beyond science or within science, the sort of human connection is so incredibly valuable and sustaining for that relationship and sustaining for me professionally because the sort of old 1960s doctor of like they're there and you know sort of walking out of the room and being completely emotionally detached. But there were some great doctors in the 1960s. There were. But you know, but being completely emotionally detached is not the way I choose to practice. And you know, we talk a lot about physician wellness now in the medical school curriculum. So with my dean hat on, often my days are dealing with students and their issues of you know, trauma on the wards and what that stuff looks like. And what I remind students is that it is not simply an exchange of knowledge. It has to be a demonstration of care and a demonstration of humanity in order for the relationship to be mutually beneficial. So I understand that a year ago, you two gave a talk about how oncologists and radiologists can better communicate with each other. And so Meg, you're a radiologist, Sam, you're an oncologist. Can you share a bit about what makes really effective communication between oncologists and radiologists even when they're not married? Where do you two are? Yeah, I think, you know, Sam and I have kind of alluded to. We see a lot of the same patients, but in a slightly different way. So Sam will see them, you know, talk with them about their diagnosis. And then a lot of times I may see them kind of from the inside, you know, on CT. And so I'm not, I maybe see their picture, but they're not standing in front of me when I read their images. And so a lot of the time what I'm saying is exactly objectively what I'm seeing and I'm saying it for the referring provider. But now with the 21st Century CARES Act, a lot of reports now are released to patients. So, you know, we've all experienced this as patients ourselves, you know, in my chart or some electronic platform used to be that they maybe just get the impression of the report, which is kind of like the synopsis of the report. Now they see the whole thing. And so now I'm not only sharing my words with potentially Sam who knows what a lot of the words I might be saying mean, but also with the patient. And so it's really interesting to get feedback from Sam because he will then see the patient who has seen the report and hear what the patient has to say about it. And so it's been very educational for me. And I think we all as radiologists have had to sort of adjust our mindset a little bit now that we know that these words are going out to the patient as well. And so things that I often say that I don't even think twice about. So something is normal. I'll just say sometimes it's unremarkable, which, you know, the prostate is unremarkable. And the patient will be like, there's nothing unremarkable about my prostate. You know, all I mean is there's nothing bad there, you know, like, but to them it's all the kind of like this is on a very remarkable prostate. Or it's really confusing to them. They're like, well, what does unremarkable mean? Like, should I worry about that? You know, or not to mention, you know, like some of the other words. And so Sam and I have talked about this a lot because there are words that we sometimes say that can be really scary to patients. So if there's, for example, a lot of something, like say they have a lot of tumors in their liver, a word that radiologists love is enumerable. They're to say like enumerable tumors. That's an incredibly scary word to patients. And I've had patients say that to me like, could you please give the feedback that that is a very scary word, you know, even if it's something benign like to say that there's enumerable. And the other, you know, so Sam and I have to, our whole lecture was kind of about one thing was the 21st century care act. So you know, we use a lot of voice recognition when we dictate and it's not perfect. So typos get into the report and we try to carefully edit them, but you find that when you've just said something, your mind reads it as you said it. And so sometimes you don't see the typos when you're self editing. And so, and they can be really critical typos. Like, I might say no pulmonary metastases, meaning like no lung tumors. But what the dictation may hear is new pulmonary metastases, which is a totally different thing, you know, and so catching that is really important. So we've tried to device some strategies how we can do better because, you know, those kind of errors are really concerning to the patient. You know, using a lot of words that they don't know is very confusing, making the report too short without any comments about anything. Patients want to know, like what are you seeing? And then, you know, like crazy things like, you know, we've now moved towards structure reporting where we may have a template, you know, and at a minimum, it puts like all the details about different organs in the same place in the report. So that the referral provider knows how to find it. So there's like a liver field, a spleen field, a pancreas field, et cetera. But some people for speed will pre-populate some of those with normal values, you know. So one thing in the liver is that some people have a gallbladder, some people's gallbladder has been removed. And a normal template will often say normal gallbladder. But for patients who've had their gallbladder removed, if we don't edit that, it's very alarming to them. They're like, "Were they even looking at my skin? I shouldn't have a gallbladder. Why are they talking about my gallbladder?" You know, or the same with like a uterus if they've had a hysterectomy and you say the uterus is normal or worse, you know, a female patient, you comment on the prostate or, you know, something like that. So like, with the rise of structure reporting, I think you have to be really careful if you use, and for me, I don't use any pre-populated field. I just use the headings and I put in all my own words. But, you know, some of our savvy, younger trainees are really good at creating these reports, which makes it a lot faster. But you just have to be really careful because if you make comments about stuff that the patient shouldn't have, and it's crazy because they may have, you know, like a terrible cancer and you're saying all these important things about their cancer, but if you say they have a gallbladder when they don't, that's what they're going to focus on in the report. And so I've gotten a lot of messages from referring providers like the patient's very concerned that you commented on their gallbladder, and I'm kind of like, "What about all the other things I said?" You know, like, so that kind of stuff, I get that feedback from Sam because he's seeing the patient and dealing with the fallout of the words I say. And then there's other words that are very loaded or have a very specific meaning to him that maybe I don't know. So for example, progressive disease, you know, like a lot of radiologists might say progressive disease if things just look worse. But progressive disease has a very specific meaning to Sam and a very specific meaning to the patient because we have all these tumor response criteria, standardized criteria that we use. And if I measured all the tumors and they had increased by, I don't know, 20 percent, some of diameters, that meets criteria for progressive disease by this standardized criterion. And that means that probably they need to go to a new line of therapy or maybe they don't have another line of therapy. So I need to be very careful if I say that word, that that's what I mean. Because that's what he's going to understand and that's what the patient's going to hear. And so I find myself saying all the time to trainees because that's a word that they like to kind of throw around. They'll be like, it looks worse. I'm going to say progressive disease. I'm like, don't say progression if it's not, if it's not unequivocal because that is, just say interval increase in size and number of metastatic lesions or interval worsening of disease. But don't say progression because that has a very specific meaning. And so I think understanding that is really important because some people, especially or early trainees, maybe aren't familiar with some of these response criteria that the oncologist are using, especially in the setting of a clinical trial or something. And so I think we have to be very thoughtful about the words we use. And Sam gets a lot of that feedback, sends it to me. And so I feel like I'm always telling, you know, and so I'm lucky that I am married to an oncologist because I get a lot more directed feedback, maybe than someone who's not. But I try to propagate it in the reading room. Like, please don't say that word. And of course, as an instructor, you probably know this too. Sometimes you have to say things a lot of times before it gets through. So I just keep saying it. I know that in certain capacities. Yeah. That's right. And for us, not only providing feedback on like, hey, this is what patients are interpreting on what you say, I tell my trainees to have hypotheses when they order a test. You know, like truly using the scientific method of like, why would you order this? Why would you? What is the question you're hoping to? answer with this procedure because there's cost behind it. There's risk to any intervention we do, whether it's a blood draw, a CT scan, a surgery, radiation, chemo, whatever. There are risks to the things that we do and the biggest risk, well not the biggest, let's be honest, but our risk is not knowing what you're gonna do with the results. So ordering something and asking a question that you're in no way ready to deal with the facts that you might get back. So when I order a scan and when I have trainees order scans, I'll say, "Hey, you know, put your hypothesis in here." Like, new diagnosis of pancreas cancer, looking for metastatic disease, comma, specifically look at the liver thing that they saw on the emergency room ultrasound when this guy initially showed up, you know. So treating your colleagues like consultants and asking them for a very specific answer gives them the information that they need to help give me the answer that I want. And so, you know, putting all of that together, having a hypothesis being ready for the if-then's that come along with it. And then what I often do to sort of close that communication loop is to sit with the patient with the report open and the images open. And this is why it's beneficial to be married to a radiologist because I can at least look at the pictures. And, you know, sit with the patient and say, "Okay, when she said this, this is what we're referring to, you know, so to help sort of demystify that." And also give the patient the frame of reference for, "Oh, that's what that means." You know what I mean? And the other piece of feedback that I give Sam, you know, for communicating with us, as he alluded to, there are so many ways that we can tailor the examination, you know, not only the modality and depending on what their question is, you know, we may decide if you really want to characterize, you know, really good tissue characterization, probably a mod, for a lot of locations, it's going to be your best test. If you want to, you know, get a really global look at everything or you're not really sure what's going on, probably CTs can be your best test. You know, so ask us. Tell us what your question is and ask us. But even within a modality, for example, like within CT, there's lots of ways we can tailor the protocol of the scan to answer the question. So with pancreas cancer, you know, maybe you want to know if there's metastatic disease, but maybe you also want to know if it's resectable because that's going to be, you know, the best prognostic. I mean, that's going to be the best treatment for the patient. So we really need to look at the blood vessels and are they involved with the tumor and all the stuff. So there's lots of ways we can change the way we obtain the images, even within CT, you know, to get a better answer to that question. So the more detail they can give us, the better we can sort of tailor the exam. And I think because, you know, we've worked in dark rooms for a long time, people think of us kind of as, you know, technologists or, you know, like we just read the images and whatever, but we can also be consultants. So if you don't know what the best imaging modality is, ask us because that's what we do. Or don't tell me which site you want me to biopsy, ask me, which site I think is the best to biopsy. What's going to be the safest and the highest yield and also stage the patient's disease. And I'll also tell you what imaging guidance we should use, you know, so that treat me as you would any other consultant, even though you think of me as like this, you know, image reader in a dark room, you know, like this is our area of expertise, you know, so ask us, you know, like leveraging, leveraging your expertise. Right. As you would any other consultant, you know, it's just we're very technical sort of in our expertise. So I think people just want to tell us what to do and then have us do it. But this is what we do. So ask us because we can we can really answer your question. We can't always, you know, sometimes there's not a great study. But sometimes if we know what the question is, we can really, you know, triage the patient to the right exam or, you know, sample what we think is going to be the safest highest yield site for whatever it is they want to know, not only to establish a diagnosis, but maybe they also want to know, are there some genetic mutations associated with this tumor that we could target, you know, like we need to get viable tissue for that. So let us look at the images and decide, you know, what site we think is going to be best for that, etc. So I want to move on to talk about how some of this relates to high school science. And by high school science, I also mean, you know, more advanced middle school science and also some of the science early college students are taking kind of like the basic chemistry, physics, biology, curricula. So Meg, I thought we could start with imaging. I'd love to talk about, you know, the basics of CT scans, although if another scan comes to mind in this question, that's okay too, a different modality. But you know, students are learning about the electromagnetic spectrum and basically all the science classes. So how do, you know, certain kinds of radiation help create these really powerful images? And maybe you could pick one kind of radiation, it could be x-rays, it could be something else and just talk about how it's useful. Yeah, I mean, it's really, I just, all of this of course was developed before I became a radiologist, but like it's just really brilliant application of physics. And so for example, CT, which computer tomography, which I consider, you know, kind of our workhorse imaging modality, I don't know if you've ever seen a CT scanner, but essentially it's a, it's a moving table that goes into a gantry that has a tube with voltage and current essentially. And it generates a beam of ionizing radiation that moves around the patient. So it's like a helical scan. So it's circling the patient, directing this radiation at the patient. And what happens is the radiation goes through the patient and it's picked up by a detector on the backside. And there's lots of different kinds of energy that's generated as a beam goes through the patient. But at the simplest level, what happens is different tissues attenuate or block the radiation in a slightly different way or interact with the energy in a slightly different way. And it creates, you know, this image of all these different gray levels. So there's something like 256 or 256 gray levels in the image. And you know, my eye probably can't detect every single one, but they also get assigned a numerical value based on the tube voltage and the current that you use. So different types of tissue or different types components have a different number assigned to them. And we measure them in how it's filled units on CT. So for example, if you take the liver, most of the liver prankama is sort of a soft tissue, you know, type tissue. And so if I put a region of interest on it, I can draw a little circle on it. It will measure essentially how dense it is or how it's interacting with the energy and give me a number. And you know, I can take the tube voltage and kind of extrapolate, well, what kind of tissue must that be? Whereas if I look at the gallbladder, which is filled with fluid, it's going to look really different on CT because that fluid is going to interact differently with the energy. And so if I would draw a region of interest on fluid, I know that fluid should measure somewhere between zero and twenty-hounds field units. And that's true of other fluid filled structures. So sometimes you might have a lesion or a tumor in the liver and you don't know what it is. You want to characterize it. Is this something benign? Is this something malignant? Because of the way it interacts with the ionizing radiation, I can draw a region of interest on it. I can measure how dense is it and I can sort of make an estimation. And sometimes it's hard to tell. There's some overlap. But if it's a fluid filled structure like a cyst, which is benign, then it should be pretty low in attenuation. It should be, you know, fifteen-hounds field units. And it has a certain appearance, you know, it should be well circumscribed, etc. So I can say, you know what? This is benign cyst. I'm not worried about this. But if it looks different, if it measures higher in attenuation, you know, then I might worry. And then there's other tissues. So I kind of say sometimes that radiology or CTE is kind of like really low power microscopy. So like fat should have a certain attenuation on CTE as a very characteristic look. And so if I see something that contains macroscopic fat, it gives me a very specific idea, depending on where the fat is, about what it might be. So if I see a fat containing lesion in the kidney, there's a very specific differential diagnosis for that. I see a fat containing lesion in the liver. Same with air. So air has a very specific appearance on CTE, really, really, really low negative hounds field units. So some places you're supposed to have air. So like in your colon, you're supposed to have air. But when you see air in places you're not supposed to see it, that can mean there's something bad going on, like a gas-forming infection. Bone or calcium has a very specific appearance on CTE. And so if I see a calcified lesion, you know, and I can be confident that it's calcium on CTE because of how it interacts with the ionizing radiation, that shapes my differential diagnosis. So to a certain extent, I think of radiology as kind of like low power micecrastic be because of how the tissues are going to interact with that ionizing radiation. And I'm not explaining the physics in great detail very effectively, but that's kind of what I'm thinking when I'm looking at images. And then sometimes, you know, if you don't have a very specific tissue type that you can identify, you know, then you kind of give a differential diagnosis. Here's where it's located. Here's kind of what it looks like. Here's three things it could be. And here's maybe what we might do now. next to figure it out. Maybe we could do another imaging modality. Maybe if we use MR, we'll be able to characterize that tissue a little better. Maybe we actually need to sample it. And this would be amenable to Armstrong-Guy-A-Biopsy or something like that. So that's kind of how what we're thinking. And I love radiologic pathologic correlation. So when I look at something and I say, OK, here's my low power of my cross-view. This is kind of what I think this is. Then maybe it goes to biopsy or surgery. Following up with the pathologist, and saying, what did this look like to you? And even seeing the cells sometimes is really helpful for me to understand why it looks the way it does on imaging. And so that's a sort of subfield of radiology and pathology that I love because I feel like it makes you a much better radiologist if you understand sort of the pathologic basis of disease. For example, if you have a fibrous tumor, this really kind of like fibrous scar tissue, it sometimes has a very specific look on MRI, for example. So because of how the cells are so tightly packed together and you've got all this fibrous stroma, has a very characteristic appearance on MRI. And you can see why when you look at the pathology, there's all these collagen fibers and all these sort of spindle cells. And so that's kind of what we're doing. And we can't always pin it down, but we're trying to use the information that we have. And sometimes it's kind of gestalt. And sometimes it's very quantitative, depending on the imaging modality used to kind of narrow your differential diagnosis. And Sam, I would like to talk about targeted therapy of tumors based on the genetic profiling. This really gets at the heart of what students learn in biology about genetic mutations. They happen sometimes with profound consequences. But also seeing them can kind of lead us to dealing with those consequences, developing therapies that get right at them. I think when Meg describes radiology as low-powered microscopy, I think the complete flip of that is really using basic science principles that A's the C's, the T's, the G's of DNA to see not even microscopically, but sub-microscopically about what's driving the cancer cells. And so an understanding of how protein conformational folding happens and all of that stuff has been such a breakthrough in the last 15 years of my practice. I've actually had to go back to some of the organic chemistry and biochemistry that I learned in college of like, OK, how does that happen? And how can we then capitalize on it? And I think that's the stuff that's really cool. A very specific and nerdy example right now is there's a specific-- when I teach students about cancer biology, I often tell them there's this two-hit hypothesis, right? There's a driver mutation that says divide more quickly or grow faster. And then there's a tumor suppressor loss. So it's like a stuck-gas pedal and a cut-break line. And you need one of those things is bad, but both of those things leads to oncogenesis leads to cancer development. And one of the cool things that biologically is happening right now, there is a mutation in a gene called K-RAS, which is common in a ton of the cancers that I treat, colon cancer, pancreas cancer, biodectoc cancer. And for a while, it was thought to be undruggable. But some really smart folks got together and saw that hey, for a specific pocket of how the K-RAS protein is folded, can we stick something in there a drug to inactivate the K-RAS growth signal from being sent down? And that was such a breakthrough that now there are sort of pan-RAS inhibitors, where it's not just a very specific key in a lock, but maybe we can get like a skeleton key to figure out how to slow growth in any K-RAS mutant tumor. And that's some of the super exciting stuff that really harkens back to a basic understanding of how do the A's, the C's, the T's, and the G's turn into hair and skin and guts and pancreas is. So it is really fun to see all that stuff develop and think about how can we capitalize on what we know. Along those lines, now that we know that that's what they want to know, at first we were like, well, how can we sample these tumors better to get that information? We know we need more tissue and we need viable cells. So we've started to kind of change or add things to how we sample tissue to do a better job to identify that stuff. But that along the lines of red path correlation, are there imaging signatures to some of these genetic changes? Can we do a better job? Is there more data that we can extract from our images to say, maybe there is a K-RAS mutation here? What do K-RAS mutations look like? And so that's kind of where we're trying to find work backwards. Like when we know the K-RAS mutation is there, we go back, we say, what did it look like? Does it look different in some way than a wild type? And maybe I can't see it with just the naked eye. I look at it and I'm like, oh, it looks the same to me. But maybe the radiomic profile is different. And so we're trying to find ways that we can either invasively or even better non-vasively better characterized tumors for them so they can decide how they're treated. So we have a listener question. This is from Emily in Phoenix, Maryland. Emily is a nurse practitioner. And she asks, I'm curious about your thoughts on the use of AI in cancer detection. And of course, Meg, I did see that your research was featured on the cover of a journal called Radiology Colon Artificial Intelligence. This is a broad question. I would be curious on some of your top of mind thoughts on this question. Yeah, I mean, this is a really interesting question. And actually, we both get asked this question a lot by learners who are trying to decide where they want to go with their career. In 2016, there was a very prominent AI researcher who basically said, we should stop training radiologists now, because AI is going to replace radiologists. And so for a lot of students, I feel like they were very nervous about choosing radiology as a career because they thought we're not going to be needed anymore. And definitely, AI, as everyone knows, is moving at a very fast pace and has a lot of potential in terms of what it can do. Right now, I would say-- and there's so many places in our practice that we could use it. The place we use it right now is for-- I think everyone is facing growing volumes in their radiology practice. I think part of it is population growth. Part of it is closure of regional medical centers. Some more patients are having to go to a smaller number of centers, aging population, et cetera. So our volumes are really off the charts. And we feel like we're reading as fast as we can all the time. So sometimes there may be an emergent finding somewhere in my list, but it may be a while before I get down to that scan. And so we have some AI modules that are tuned to a few emergent conditions. Like, oh, do they have a bleed in their brain? Do they have a clot in their lungs? Do they have a collapse of their lung? Do they have a perforated bowel loop, et cetera? And it will flag it for me. So it might move it up the list for me if it has some sort of emergent finding. And these algorithms are not perfect. They're not always right. So it may think it has, for example, a bleed in the brain, and maybe it's something else, but they're getting better all the time. And so that's really helpful. Number one, for helping me triage, which studies are on my list or the most urgent, is there something I need to get to first? And to function is almost like a second reader. In case I'm going fast-- oh, I have an alert here. I better look one more time at that thing, make sure that I'm not missing it. And so that's one way that we're using it. And for a lot of the applications, they're emergent. So they're not so much tumor detection. But they're helping me look at a cancer patient faster that might have some sort of complication versus one that maybe doesn't. One way in which I've seen AI in radiology and in oncology be super helpful, when you think of CAT scan, it's basically a series of two-dimensional images sort of stacked tightly on top of one another to sort of allow for rendering a three-dimensional image. And in the lungs, little blood vessels can look a lot like round tumors when cut on edge. And so there is something that has happened with our radiology software in the last couple years, where AI is able to subtract out all of the blood vessels of the lungs and only leave behind with a pretty high degree of precision stuff that's in the lungs, which is really cool from a cancer detection and from a therapy monitoring standpoint that would have been science fiction to me 10 years ago to say like there's no way you can digitally subtract those blood vessels. I would love to be able to do that, but I can't right now. And now with the advance of better computing technology we're able to do that. Yeah. And even quantifying tumor size, for example, it takes a long time to manually measure stuff. And when you get artificial intelligence-based tools that can identify and segment and measure a tumor for you, it's going to give you a probably more accurate reproducible measurement depending on the algorithm. The longing is a place you can imagine it's an air-filled structure. So when there's a tumor that's soft tissue, it's a really high contrast task, meaning the air around the tumor looks very different than the tumor. The liver can be a little trickier organ because sometimes a tumor is really close in a tenuation or density to the background liver and so those can be harder to detect and segment. So definitely, you know, those are really useful tools. The article that was on the cover of our artificial intelligence, it's all about these automated tools that we can run in the background. So one thing we've been really interested in is we're getting CTs for a variety of reasons. Maybe they have abdominal pain, maybe they have cancer, but there's a lot of data there, meaning like we could look at their visceral fat and everybody has a different amount of visceral fat. We can look at their subcutaneous fat and the amount of each that they have. We can look at how much they are at a calcification they have, which might be an indicator of cardiovascular disease. We can look at liver fat, we can look at muscle fat or muscle bulk. You know, and so all of these things can be indicators of frailty or indicators that they may have an elevated cardiovascular risk. And so we have some clinical risk scores that are pretty well known, like the Framingham risk score for example, where you take a patient's age, you know, there's sex, some of their laboratory values, etc. And you can predict their 10-year risk for having a heart attack. If you compare some of these CT imaging biomarkers, like visceral fat, aerodic calcification, liver fat, muscle fat, we can actually do a much better job of predicting their cardiovascular risk. And we call it sort of opportunistic screening. And a lot of these tools at first were manual. And so it's very labor intensive to have to segment out the visceral fat. But when you use artificial intelligence, you can totally automate these tools to be running in the background, which we now have. And so largely they're still in the research realm because it's difficult sometimes with some of these AI tools to train them appropriately or to understand what factors might indicate or what might impact normal values. So for example, you know, we've now run these tools on hundreds of thousands of CT scans because in a given year, you know, we may do a ton of scans. And we can sort of establish some normal values within our patient population. But our patient population may not mirror the global population and there may be factors like geographic location or race or age that influence what is normal and what is not. And so these biomarkers are not totally ready for, you know, population-based screening until we can figure out those kind of questions. And so right now we're running these AI-based tools in the background in the research setting in all different patient populations because that's another issue. You just need to make sure that your AI tools are well trained and don't have inherent bias based on whatever your patient population is when you try to apply it to another patient population. And so that's a really big thing. So that's one potential application of AI. You know, it's gotten very good at, you know, for example, looking at a chest X-ray, which is a single 2D image. There's still a lot of work being done on a CT scan because, as Sam said, you know, you're this series of slices, lots of organs, lots of processes, but it's getting better all the time. So, you know, I hope that it can make us more efficient. I hope it can, make, help us triage, you know, high risk patients. I hope it can automate some of the tasks like measurement to make them more reproducible, accurate and faster. I hope we can, you know, be pulling more data from our scans, just ringing as much data as possible from the imaging that we're getting anyway. I hope it won't totally replace me. I hope we can use it in that way rather than, you know, me losing my job, but I don't, I don't totally know. It's changing really fast. So, I think, and then there's this whole idea around, like, large language models, you know, and they could generate a whole report for me, and I could edit it, you know, and it's still not at a place where maybe it would say exactly what I would say, or we can do the reverse, you know, like visual grounding where I say some words, and Sam's trying to show the patient what I'm talking about, large language model, can sometimes take my words and show on the image what I'm talking about. And, but again, it takes a lot of scans, a lot of different indications to train for that. That's great for learners, too, because maybe they'll read a report and they're like, I don't totally know what she's talking about, but if we can do visual rounding, here's the phrase she said, here's what she's talking about. It's a great educational tool, too. So, there's lots of really exciting AI applications that's just there are bearing stages of development, and it's been a very active area of research. So, what advice would you give to students who are interested in science? Oh, man. I think, for students interested in science, it is such a great time to be excited, right? There's so much great stuff that's happening, and recognizing that we're standing on the shoulders of giants, right? Like, there are so many advancements that in the moment can feel incrementally small, but are when looking backwards are so great, you know, thinking about the human genome project, right? Like, sequencing the genome was infeasible 25 years ago, and now, hey, I can get it back in a week, you know? So, staying true to the scientific method of having a hypothesis, testing it, and critically analyzing the results, especially when they're not what you expect, is the kind of thing that really makes a good scientist. And I've in my years working in science, I've learned that so much of it starts with asking the right question, and then being honest in your analysis, you know? So, if I had any encouragement, or as I encourage young learners to go into science or medicine, it is keep asking questions, keep digging a little bit deeper and trying to push a little bit further with the understanding that you're not going to answer all the questions at once, but if you answer even a little fraction of the question that you have in your head, you are moving the ball, right? Like, you are moving science forward. And so, I would, you know, as we talk to our own kids, just keep asking questions, keep-- Be curious. Be curious, not judgmental, back to Ted Lasso. And so, thinking about that curiosity and cultivating that. And I guess I would say, you know, not to get discouraged. I feel like sometimes right now, it feels like people don't value science in the way that maybe they once did, you know, funding sources are much more challenging. Sciences, I consider it, you know, rigorously tested and unquestionably factual. I mean, it's a little bit more of an art than that sometimes, but I just-- I hope that this is, you know, just a cyclical trend that's going to kind of have an flow and that things will look up again. So, I just don't want people to give up on science, or get discouraged about science because I think, as Sam said, it's so incredibly valuable. Not only in its purest form, but also just the general approach to problem solving. You know, I think it's really important to embrace that mindset, to ask questions, think about how you can answer them, and problem solve, honestly, the results, you know. But I just-- I hope that people don't move away from science. I hope they don't get discouraged about science. I think it's just the best. And I hope that, you know, things will kind of, you know, cycle back in the near future. That's such a wonderful note to end on. Thank you both so much. This has been so great. Well, thank you so much for having us. We really appreciate your time, Sue. That's super fun. Great to see you. That was an interview with Sam and Meg Lubner, both working in Madison, Wisconsin, Sam as an oncologist and Meg as a radiologist. I especially appreciated their thoughts on AI and how it's changing cancer care. Listeners, please consider filling out a survey so we can continue to bring you great content. You can find a link to the survey in the show notes of this podcast. And you can also find it on the Instagram page, the account is @ScienceFairPodcast. Thank you for tuning in to today's episode of Science Fair. Please rate and review the episode on the podcast app of your choice. See you next time. [BLANK_AUDIO]