The Defense Department’s Joint Artificial Intelligence Center (JAIC) is joining forces with the Defense Innovation Unit to focus on a universal problem: Cancer – in particular, metastatic breast cancer. The work brings together AI, digital imaging, and a unique piece of equipment. For more, the JAIC’s warfighter health mission chief, Navy Capt. Hassan Tetteh, MD, spoke to Federal Drive with Tom Temin.
Insight by Splunk: Explore how data is the glue that will hold JADC2 together by downloading this exclusive ebook
Tom Temin: Captain Tetteh, good to have you on.
Capt. Hassan Tetteh, MD: It’s a pleasure to be with you and your audience Tom. Thank you for the invitation.
Tom Temin: And we should point out you are also a physician and surgeon. So you’re kind of coming at this from a couple of different ways. And let’s begin by the health mission initiative within the JAIC, tell us what that’s all about.
Capt. Hassan Tetteh, MD: Sure, absolutely. Super passionate and excited to be leading this mission in the JAIC, the Joint Artificial Intelligence Center and the Warfighter Health Mission. As you mentioned and pointed out one of the actually very important attributes about the JAIC is that we have practitioners that are leading our mission. So while I am a physician and a practitioner, still active surgeon, what I have is also some background in informatics and in AI. And that helps to bridge the gap between the practitioners that are going to be using the technology and also the technologists and the computer scientists and AI engineers that build a technology. And what this does is it affords a seamless and smooth transition for the adoption as well as the validation of the technology when we are implementing it in real time. So my colleagues in cyber and the other missions in warfighting, etc. — those colleagues of mine are also very experienced, seasoned officers that have been dedicated to their craft, but also now bring the context and the lens of how AI can be applied to the problem sets that they have and offer solutions to those that are prosecuting and moving out and our mission. That’s one of the unique value propositions of the JAIC, and one of the reasons why I was immediately attracted to the model, the design and the mission.
Tom Temin: And you’re focusing in this project on metastatic breast cancer, and a sort of super microscope. Tell us more about the project itself.
Capt. Hassan Tetteh, MD: Sure, Tom. That’s a really good point that you brought out that we’re focused on metastatic breast cancer, but I would just enlarge that aperture a bit and say that that is just the tip of the iceberg. What we’re really aiming at in terms of our mission is to leverage AI to fundamentally transform the way that we both diagnose, treat and care for our military service members, and ideally, have this impact even be translated and felt in the largest civilian sector. So metastatic breast cancer is just the beginning, quite frankly. And the reason why we chose and elected to begin with this project is because we’ve already validated this as a method and a means that works with peer reviewed publications, and with research that we’ve been conducting with our colleagues at Navy medicine, San Diego Health Center, with our colleagues from the Defense Innovation Unit, and with the actual practitioners and pathology experts that are leaders in the field. And so for many years now, we’ve actually been working on developing this technology. And metastatic breast cancer was one of the early cancers that we validated, as well as prostate cancer. And so what this mission is aimed at doing is continuing in that effort for these particular cancer types, but also expanding our technology and our diagnostic capabilities to include other cancers that will also ultimately be helpful in the workflow of the pathologists. The microscopes that you identified and talked about are called augmented reality microscopes. And if you can think about the typical microscope that all of our pathologists and health professionals use on a daily basis to diagnose a disease and to validate when people may have some conditions, sort of a microscopic level, what we’ve done is interface, placed an interface within the aperture of the lens, and using both micro display and compute unit, and a combination of a camera and computer vision — this interface allows processing to occur so that when the pathologist looks into the microscope, just like they do on an ordinary basis, but with this interface, the little rings are drawn around the abnormal tissue and the cells that are likely cancer. And what this does is it expedites the accuracy, as well as the efficiency of the pathologist. So it’s very exciting. And I think it’s going to be one of these tools that very much like the stethoscope of many centuries ago, completely changed the way that we were able to diagnose and treat the patient.
Tom Temin: And we should point out, as we said in the beginning, the Defense Innovation Unit is also joining with you on this project. What are they bringing to it?
Capt. Hassan Tetteh, MD: Yeah, that’s a very good question. I would add that not only the Defense Innovation Unit bringing their expertise and their domain knowledge experts, and of course their networks of the leaders in this field of computer vision, cloud based computing and AI, helping to make those introductions and bridge those connections between the technology practitioners and ourselves at the JAIC. They’re bringing a lot to the table. But they’re one of many other collaborators, including, of course, the practitioners, the pathologists, we’ve had several pathology champions at military treatment facilities that not only are embracing the technology, but are helping us to validate and sort of spread the technology to other practitioners. They’re also helping us with our research and collecting information and data, and in the testing and evaluation of the technology, and DIU is right along side by side with us as partners in terms of gathering all that information. As well as planning for the expansion of the capability to other cancer diagnostic capability.
Tom Temin: And you described earlier the microscope, the augmented reality microscope, and how that enables the viewer to see more than they could otherwise. Where does artificial intelligence come into this in terms of predictive analysis?
Capt. Hassan Tetteh, MD: Very good question. Excellent question. As I mentioned before, the actual microscope itself has several components that are sort of added on to the existing platform of a microscope in that interface, sort of referred to as the augmented reality component of it, has within it a micro display, and a visualization overlay, that allows the sort of ring to be visualized by the person that’s looking through the lens. But what is also included within that interface is a compute unit. In that compute unit is the intelligent processor, if you will, that has been trained on many different tumor possibilities. And in this case, as I mentioned, we’ve validated those models on both metastatic breast cancer and prostate cancer. So that’s where the actual AI is, that’s where the sort of brain of the apparatus is located. And as we develop more of these models, and do more and additional training, those algorithms and that sort of compute unit is going to be updated with new models and a new capability to diagnose additional cancers. So it’s a very sophisticated piece of equipment, but it’s one that is absolutely using AI as a point of care. In this case, it’s a point of diagnosis for facilitating, again, the accuracy and the efficiency of the pathologist so that we can improve the speed, the diagnosis and the accuracy of diagnosis — and ultimately help to care for our patients. And in many ways that helps save lives.
Tom Temin: There are many inputs then to this project, the microscopes, the intelligence, the compute power, the AI, the algorithms, what do you see as the output is it a new process, is it a new diagnostic tool. What will be the deliverable as you envision it?
Capt. Hassan Tetteh, MD: Yeah, very good question. Let me just add one other very critical input, in addition to all the ones that you aforementioned, and that critical input is data. And I think that is where the military is going to be the vanguard. Our data if you think about the way that we capture and collect and retain over time, the data of servicemembers record for potentially in terms of imaging and in term of the inputs from health practitioners with their various visits, that data unlike any other place anywhere else in the civilian sector at large. And within the military, we have a comprehensive collection of data, and its volume, its variety, and strategy is much richer than it would be in practice civilian sector, just because of the demographic makeup of military service members and the very unique diversity that we have. What that does is it makes it very rich for training material, which is what you need for anyone that’s working in AI, they know that data is sort of the resource that you need, and our data happens to be very good. And so what that does is it helps to create an output, if you will, it’s of high fidelity and is very great utility. So for the output is what you asked in terms of what will be the output, the output will be the very beginning of transforming the processes by which our practitioners are using technology to help enhance their diagnostic capability. I used the stethoscope as an example, way back when. If you think about how that piece of equipment and that technology back in its era, fundamentally changed the way that practitioners were able to assess the condition of the heart. In this case, it created a leapfrog technology back then, and even now continues to do so in terms of helping practitioners assess more accurately what is happening with the patient. And when they get the accurate diagnosis and a better diagnosis, they’re able to affect the better care plan. So in the case of the augmented reality microscopes, again, this is the beginning of helping to increase the efficiency and the effectiveness of the diagnostic capability of our pathologists, beginning with these two cancers, but expanding over many other conditions and other cancers and pathology. And what that will do is help to create a more effective and streamlined diagnostic capability for our practitioners. But ultimately what we’re going to be doing is improving the accuracy and availability of an expert diagnosis, and that is something that will help to serve our servicemembers and patients in a very good way, in a very impactful way, and in a way that helps save lives. Finally, just to kind of accentuate that, the output here at the end of the day is to help save lives and help improve the care that’s delivered to our patients.
Tom Temin: You might say, peering into that microscope then you’re really looking into the future.
Capt. Hassan Tetteh, MD: Oh, absolutely. And if you think about, again, the rich and unique data set that we have collected throughout several decades in the military, this is a very rich data set that I think is going to help to inform a great number of these AI technologies that are now being developed so that pathology and radiology — and also with natural language processing and the health record as well. So it’s a very exciting time. We’re at the cutting edge, I think the military is doing what it’s always done. We’re being leaders. If you think back to Washington’s Continental Army and the introduction of mass vaccination, in that case, it was a better solution. But that led to the developments in sort of the propagation of vaccines that has obviously infiltrated into the civilian world and sector and helped a great number of people. It was also in the Army back in the Civil War that introduced the utility and the use of actual blood transfusions. And then the Army, to their credit, their great credit, helped to pioneer the actual safe and effective banking of blood. These great technologies not only serve the military well at the time, but of course as now are permeated into the civilian sector and has actually contributed greatly to modern healthcare and it’s a great reward that we are benefiting from those advances. And I think that AI is sort of the new technology of the day and the military and our partners within the military and the DIU and our practitioners — and of course, those of us here at the JAIC — are all excited to be a part of this great new dawn if you will.
Tom Temin: Navy Captain and Dr. Hassan Tetteh is the Warfighter Health Mission Chief at the Joint Artificial Intelligence Center. Thanks so much for joining me.
Capt. Hassan Tetteh, MD: Oh, thank you, Tom. It was a pleasure and great delight to share our great work with with you and your audience — and we look forward to having an opportunity to continue to update you on the progress that we’re making.