Interprofessional Synergy: Transforming Radiology with MR Fingerprinting

Click the play button to listen now:

Subscribe: Apple Podcasts | Amazon Music | Spotify | iHeart Radio

Daniel Simon, MD: Hello everyone. Thank you for listening to another episode of Science@UH I am your host Dr. Dan Simon and today I am happy to be joined by Dr. Chaitra Badve and Dr. Dan Ma.

Dr. Badve is the Director of Magnetic Resonance Imaging at UH Cleveland Medical Center and Associate Professor at Case Western Reserve School of Medicine. Dr. Ma is an Associate Professor in the Department of Biomedical Engineering and Radiology at Case Western. They are the part of the radiology team who are at the forefront of the revolution of magnetic resonance fingerprinting imaging. Welcome Dr. Badve and Dr. Ma.

Chaitra Badve, MD:  Thank you for having us.

Dan Ma, PhD: Thank you, Dan Simon, for having us.

Daniel Simon, MD: Great. Well, it's great to be joined by both of you today. And I think before we begin, I want to take one little step back for some of our listeners who may not be familiar with principles of imaging, both CT based and MRI based. Maybe we could get really basic, to start out, and just ask you to tell us the difference between radiation based CT imaging and MRI imaging. What are we measuring? And what is the big advance of MR first over CT? Maybe Dr. Badve you could start with that.

Chaitra Badve, MD:  So CT scan as we know is a radiation based technology that uses X-rays that pass through a patient and based on the tissue density, how dense the tissue is, it is going to create an image which is density based. So, when you have a CT scan of head, you have certain Hounsfield units or density measures for each structure in the brain that will be a different measurement for CSF, for gray matter, for white matter, for skull, air and so on. And water is considered as Hounsfield units of zero and all the other structures are based on comparative density to water. Magnetic resonance imaging, on the other hand, is a radiation free technique. It uses very strong magnetic fields that can stimulate the water molecules in human body or human tissue and based on the signal that is generated by those protons an image is generated.

So the strengths of MRI are; there is very superior soft tissue resolution and that allows us to take a detailed look at structures within the human body, which is not afforded by CT scan.

Daniel Simon, MD: Great. OK, so now here's the big question for both of you. I've had a lot of MRI scans in my lifetime, and I've always wanted to know what's all the banging about? Why is it so noisy? So, Dr. Ma, tell me, why is it so noisy?

Dan Ma, MD: So the MRI scan is a big magnet. The noisy sound is from the actual scan. So whenever there is a scan, you will hear the loud noise. Of course, there's some technology to reduce the noise, so typically we have, earplugs and headphones and there's also in terms of research, we also develop some technology. There's also other product available to reduce the sound, so the sound, it's from the scan itself. It's really hard to reduce the sound. That's why we generate image from those. But there are technologies to reduce the sound.

Daniel Simon, MD: Great. OK. So now we're gonna move into this new part of MR Imaging, which I guess you would say to a layperson like me, a cardiologist, you know, we're just plumbers. We're not as sophisticated as you, imagers. Is the ability now to do special things to these images, a combination I guess, of image analysis and AI and other things to learn a lot of new things about tissues, almost like virtual histology? Perhaps you could explain to us, Dr. Ma, you were on the first author of this Seminole paper in nature about MRI fingerprinting. What is it exactly? Why is it so powerful?

Dan Ma, PhD: To start with, MR Fingerprinting is a new scan, so it's a different acquisition and also involve different image analysis. So we generate different images as compared to clinical MR. The big difference is that the new images, we call it quantitative measurements versus most current MR scans, which provide qualitative measurement. What does that mean is that, so, imagine MRI as looking at a picture where you can see which areas are lighter or darker that would indicate possible lesions or disease but a bit like checking whether someone has a fever by feeling their forehead, so it's a feeling, or you can read it by yourself, but there's no exact number and MR fingerprinting on the other side is like using thermometer to get the exact temperature from that person. So now you have a number that indicate how bad of the fever is, so this number will give doctors a specific number or tissue measurements that can help to spot the disease earlier or understand the disease better.

Daniel Simon, MD: So as a cardiologist who looks at an MR of the brain, obviously, as you pointed out, there are areas that are bright in areas that are dark. And what you're telling us now is that MR fingerprinting is going to tell us new information. And maybe Dr. Badve, you could help us understand what is that information telling us? Is it telling us the difference between normal tissue and tissue that has a brain tumor? Is it telling us something about vascular structures? What does the MR fingerprint give us insights into?

Chaitra Badve, MD:  So MRI fingerprinting is best defined as a framework. So you can change the sequence to make it sensitive to different tissue properties, T1 and T2 relaxation times, which are the fundamental units of MRI's, are the most basic things that we are measuring. But you can make it sensitive to measure other things like that…products are slow. Based on MRI and MR fingerprinting information, we are able to define ranges for the first time for normal and abnormal. So, as you said, you can differentiate normal tissue from abnormal tissue, but even in abnormalities you can differentiate or characterize the lesions further and classify them. So for example, in our work we are showing that you can differentiate between grades of tumor, you can differentiate between IDH mutant and wild type tumor. So, it gives you information that is above and beyond routine MRI and most of this information is actually actionable, which can contribute to patient’s direct care.

Daniel Simon, MD: You're actually telling us, in essence, that an MR fingerprint is almost like virtual histology. You can look at an image, you're not really biopsying the tissue, but you can actually tell the characteristics of what kind of tumor this is. Is it gonna be aggressive? Is it responding to treatment? Is that what you're saying?

Chaitra Badve, MD: Yes, exactly. MR fingerprinting is, for the first time giving us that information. You have the ability to see some of this on MRI, but as Dan mentioned, Dr. Ma, that this information is very qualitative and it depends on the reader skills, the scan quality and all of that. So. It's usually defined as hypointense or hyperintense. And now we are moving away from subjective interpretation to being more objective and data-driven in making these calls.

Dan Ma, PhD: I would like to add two things from what Dr. Badve has mentioned. So, I see another benefit of having those numbers is that you can now detect some mild subtle lesions now and it can improve the sensitivity of detecting those disease earlier.

Let's take this fever for example. It is really hard to tell, low fever, of course, if their forehead is super-hot, then you can say they’ve got a bad fever, but what if they only have a, you know, mild fever? They may feel sick, but not necessarily high fever. So in this range, I think giving a number is helpful.

So, for example, we've applied MR fingerprinting to detect very subtle lesions from epilepsy patients where typical MRI cannot detect. So this is one example and another factor, I think, having number is useful is if we look at longitudinally. So we want to track the change.

So for example, we have an application of using MR fingerprinting for pediatric patients. We will want to measure, what's their developmental delay? So we measure multiple points along their development and using those numbers, we can track their developmental trajectories and predict whether they have developmental delay or not. So, having those number or measuring the difference can really help us on this rather than saying whether their image is brighter or darker.

Daniel Simon, MD: Thank you very much. That's really helpful. So congratulations to both of you on your recently awarded $3.5 million NIH grant to study MR fingerprinting and analysis platform for brain tumors. This grant which you both serve as the principal investigators is a great demonstration of interprofessional team science collaboration between basic and clinical scientists. Can you tell us more about the project? What are you hoping to accomplish?

Dan Ma, PhD: So, this is a multiple site academic industry partnership grant that involves Case Western Reserve University, University Hospitals, University of Pennsylvania and the Siemens Healthineers. So this collaboration aims to integrate advanced imaging technology such as MR fingerprinting and AI based image analysis into everyday clinical workflow. So ultimately we hope to benefit patients worldwide.

So from the technical side, one of our goal is to enhance the speed and accuracy of brain tumor diagnosis through new MR Scan and AI based prediction method. So those technology, what we want to be easily adaptable in the clinical setting. So we are tackling technology challenges to ensure that this innovations, this technology, is not only meet the high standard for different patient needs, but also seamlessly fit in within very busy schedule of the clinical environment.

Another key focus from the technical side is transferring those new technology from a research tool into a practical clinical product. So that involves developing some infrastructure in collaboration with University Hospitals and Siemens Healthineers. So this involves translating MR fingerprinting scan from a research tool to an FDA approved product so that the scan can be installed in any MR scanners and it can be done as a, press the button, easy to operate.

So radiologists will be able to read those MR fingerprinting images like other clinical images without waiting for research process. This also applies to image analysis and disease predictions, which will also be available in the clinical setting, so radiologists don't need to understand the technology background in order to, you know, use those tools and interpret those results. And the last point I want to mention that this grant is academic industry partnership. So Siemens will play a big role of integrating those technology into the clinical workflow and also help us to expand this technology to a bigger impact. So let's say expanding this distribution of this technology worldwide. This is through their global digital market. They will make this technology available for all the Siemens sites over 4000 institutions in 60 countries. So this grant will aim to improve this brain tumor diagnosis tool and treatment process and also ensure that this innovations can be accessible worldwide, to have a bigger impact.

Daniel Simon, MD: Dr. Badve, when I've gone for a scan, it takes 30 plus minutes. Tell our listeners how much additional scanning time is there for an MR fingerprint, or is it part of the routine scanning? Does it add double the time or is it only a few extra minutes?

Chaitra Badve, MD: It's a few extra minutes, in short, so as I was saying earlier, MR fingerprinting can measure multiple properties and the most commonly used can measure the T1, T2. A whole brain's 3D high resolution MR fingerprinting scan of brain requires about 5 minutes for acquisition. And since it's purely quantitative data, we can use it for not just data analysis on the quantitative maps that we are using for various research purposes, but also generate synthetic images. So all the images that are routine MRI scan has T1 weighted T2 weighted layer. We can actually generate them from the single acquisition, so it is very possible in the near future to have a 5 minute scan that the patient will have. It will give you quantitative information about T1-T2, myelin, brain volumetry and plus give us all the qualitative scans that the radiologists are used at looking at. So it has multiple advantages in terms of acquisition and processing.

Daniel Simon, MD: So we've talked a lot about MR fingerprinting in the brain, but I understand that it's also now being applied to breast, pancreas and prostate as well. Can you just let our listeners know about where else it's being used and what its application could end up being?

Chaitra Badve, MD:  MR fingerprinting is being applied outside the brain, the key areas that UH is working on are breast imaging, renal cancers as well as prostate cancers. And we have wonderful radiologists here at UH, who are leading these projects. All of these projects have been funded by independent RO-1’s with federal funding and the initial results are really stunning. So for example, in breast imaging the preliminary data shows that you can differentiate between responders and non-responders for new adjuvant chemotherapy, which is a very, very important clinical question to address.

In renal cell cancers, again the focus of clinical application is to differentiate between high grade and clinically significant neoplasms from neoplasms that are not as clinically significant and the same for prostate cancer. So actually in prostate we have made a lot of progress where we have looked at various aspects of prostate cancer and differentiating lasers that can be just surveilled with passive surveillance, with lesions that need active treatment and the people who are leading these projects are Dr. Marshall in breast imaging, Dr. Tirumani for renal cell cancers and Dr. Bittencourt for prostate.

Daniel Simon, MD: How are we integrating magnetic resonance fingerprinting into clinical care at UH right now? Is it being used to make clinical decisions or is it only research based tool?

Chaitra Badve, MD: It is both, and UH, it is the first institution across the world where we have integrated MR fingerprinting scan into the clinical protocol. Currently, this is available only for brain imaging, but very soon it will be expanded out to breast, prostate imaging as well.

For every patient that's coming into CMC, the standard MRI brain protocol now includes MR fingerprinting as a routine. And these images are rapidly reconstructed and immediately shared on the clinical path. So a radiologist has access to all the routine MRI images as well as fingerprinting maps. And we are implementing these in brain tumor patients, dementia patients, epilepsy, post op follow-ups and then just routine scans follow up strokes. So we are collecting a wealth of data to look at MR fingerprinting as a population research tool, but it is also giving us novel opportunities to do bedside or reading room research where you are looking for the first time at actually one and two values for different lesions in the brain which we were not able to measure before so it has clinical implications, but also has tremendous research applications in the near future.

Daniel Simon, MD: Great. So you know our listeners today need to know that we're talking to two superstars you're part of a group that has over $49 million in funding from NIH got another $9 million in grants in 2023 alone, and more in 2024. We haven't even counted it up yet. So you have this amazing team. I mean, stars from Jeff Dirk Sunshine, Mark Griswold, Galani all these superstars, what's your secret sauce? How do you guys, you biomedical engineers, clinical MRI people? What, how do you do this? What's the secret sauce?

Chaitra Badve, MD: I think the secret sauce is we focus on two things. One is no technology should be without a direct clinical translation. Most of the physicists, not most all of them are super focused on making their technology implementable and using it for betterment of patient care. And that makes it really helpful for a physician, like me, to work in a translational partnership with these scientists. The other thing is I think very strong focus on interpersonal relationships and collegiality and all the names that you mentioned, we are standing on their shoulders today and they have been our mentors and guides and we really appreciate the groundwork that they have laid for us, that we are reaping benefits from and I think the generation with us and the future generations will continue to do that going forward.

Daniel Simon, MD: Dr. Ma?

Dan Ma, PhD: I totally agree with what Dr. Badve has talked about. We really benefit from this very long term relationship between MD and PhD's and I want to emphasize that it takes a long time and effort to really build this environment. The first thing I can think of is the working environment, we work in the hospital, right next to the scanners. This allow us, for the PhD researchers to really feel this clinical setting. We see patient and we see what's the actual clinical workflow would be like and we scan patients. So this gives us a strong clinical needs. We know what the clinical importance or scientific importance is and we want to tackle that. This allow us to better understand the situation. And Dr. Badve’s office is right next to mine, similar to all other MD's office right next to a PhD's office. This allow us to have very consistent or continuous discussion about what's needed or what's the technical challenge talked about. We understand what's the technical challenge and I try to understand what's the clinical challenge and we talked almost every day about this. So this setting really enable us to work closely together to tackle what's important. And I also want to bring up that this team has a long standing collaboration with medical manufacturers, as well. So we have a long term over 30 year’s collaboration with Siemens which will allow us to quickly transfer any research technology to a clinical product. So we have patents, we have commercial product starting from decades ago.

For example, Dr. Grace was parallel imaging technology right now has become the routine MR scan built in every single scan really accelerate the scan two to three times as compared to the original one. We can see what research technology can be translated and for now, umm, our fingerprinting is also patented and licensed. Its FDA approved, so it will soon become a clinical product. Also distributed globally and become a product for other countries. So this is really academic industry and the hospital collaboration everything together.

Daniel Simon, MD: Well, how inspiring to be with two members of this special interprofessional team. I think you have the sense here of listening that the MD, PhD partnership is alive and thriving. It's so exciting to think that this MR Fingerprinting technology, which is going to spread around the world, started in the basement. So to speak, of University Hospital. Thank you so much for joining us today to learn more about research at University Hospitals. Please visit uhhospitals.org/UHresearch. Thank you so much, Dr. Badve and Dr. Ma.

Dan Ma, PhD: Thank you, Dr. Simon.

Chaitra Badve, MD: Thank you for having us. It was an honor.