Samuel Frishman, MDC - Robotic System to Improve Cancer Diagnosis | LSI USA '24

Samuel Frishman  0:00  
Sam, good afternoon, everyone. My name is Sam. I'm founder and CEO of MDC, and we're a surgical robotics company creating a new chapter in minimally invasive surgery. Cancer impacts everyone, whether directly or through family and friends, and unfortunately, over 40,000 patients die each year from misdiagnosis and delayed diagnosis, and this is a really tough number because it's largely preventable with better technology and earlier access to care at the same time, this is a large financial burden. Mistakes and mis prevention opportunities result in 200 billion in excess costs on the system. Biopsy is one of the first steps in the diagnostic process, and what we're specifically focused on are the 3 million percutaneous soft tissue biopsies that are done each year in the US. These are done with a long needle about 15 centimeters, puncturing through the skin, targeting soft tissue organs like the liver, the kidney, prostate and breast, and they're targeting small lesions about one to two centimeters in diameter. Unfortunately, diagnostic error happens in about a quarter of these 25% and a large part of that is missed targets. It's difficult to insert a needle 1015, centimeters under the skin and hit a small lesion, one to two centimeters in diameter. These procedures are guided with some form of imaging, MRI, CT or ultrasound, and we're specifically focused on MRI and CT because these provide the highest soft tissue contrast. You can most clearly see those small lesions deep under the skin. At the same time, the capabilities are highly underutilized. Physicians today mostly use MRI and CT for diagnostic and preoperative imaging rather than real time guidance during the actual interventions. What that looks like is shown in this animation. Needle is first inserted partway, and then the patient's put back in the scanner? Reimaged, make sure everything's on track, and then pulled back out for another insertion of the needle, put back in. Reimaged, make sure everything's on track, pulled out needles, inserted a little bit further. If you're going 1015, centimeters, inserting about one centimeter at a time, as you can imagine, that's a lot of in and out of the scanner, long procedures and preventable air and needle placement at a high level. MRI guided biopsies are difficult and slow, and we're making them simple and efficient. We've created a robotic device that provides real time access from the control room, which is adjacent to the actual scanner itself, and we do this with a hydraulic transmission. The physician is able to feel the first forces occurring at the output from the remote input in the control room. Building is shown in this animation. There's a boom that swings in and out of the way, so you can easily load the patient and the device goes in with the patient into the scanner. The physician, using the hydraulics, is able to teleoperate the system from the control room, adjusting and advancing the needle, all with real time MRI guidance. This creates simple, more accurate and higher throughput procedures benefiting all four major stakeholders in the hospital system. We've created several proof of concept prototypes to demonstrate this technology. At the top you see a video. It's a fully plastic ceramic arm with all actuation through hydraulics. It's one to one in position, one to one in force. Motions are mechanically teleoperated through the hydraulics from the input to the output. This would actually be pretty tricky to do, even with the traditional robotic system, but especially with something that works inside the MRI scanner. And a key point, there are no motors, no pumps. This is a fully mechanical system. Motions propagate through our custom hydraulics from the input to the output. We subsequently made a much more compact device. It's a five axis system, xy translation, needle advancement and two angles of rotation through a remote center of motion we've conducted an animal study with our initial system, showing that we can use the device to advance the needle through through the skin of a pig into its liver, all remotely from the control room, adjusting its angle along the way. One of the things we're really excited about is our ability to integrate AI through digitization of the transmission. So in addition to the mechanical teleoperation, we've added motors and sensors, and that's shown in this in this video. This is, of course, a benchtop prototype, but the user can teleoperate the output directly, but then they can even let go, and the motors and sensors will take over and replicate the motion fully autonomously. Of course, in a biopsy, you wouldn't just replicate motion. But what we're showing is that we have the technology through this transmission to do full mechanical teleoperation, power assist and even full automation, and that allows us to deploy these systems, gather data on the mechanical system, position, force data, with an image data set, and then eventually train an AI. Biopsy assistant. We also have a solution for CT guided biopsy. And CT you don't have the same MRI safety constraints, so you can use more traditional robotics. Here we have an off the shelf robot arm for bulk manipulation of the needle, and then our hydraulics for advancing the needle. Of course, that's, again, a proof of concept prototype, but this is what we're building next. We've got a cart system, an input Console, and our needle advancement aid sits on the cart with the robotic arm. So hopefully that gets you excited about what we're building. But what's our strategy to get on the market? How are we going to make money from this? First, we're going to strengthen we're going to improve current liver biopsy procedures. These are already done on the market and the way that I described previously, stepwise moving the patient in and out of the scanner. We make them faster, more accurate, better outcomes at lower cost. Then we want to expand to treatments, not just diagnosis through biopsy, but also ablation, and ultimately, leveraging the get data that we gather, creating our AI biopsy assistant, we want to become the gold standard across MRI and CT for biopsy and ablation. Liver cancer is a big market. There's over 350,000 procedures each year in the US, over 10,000 ablations across the different organs. We can target over 3 million procedures annually. Bottom up analysis, based on our sales points, about 3 billion in equipment sales, and then a little over 2 billion in recurring revenue from a disposable kit, software and maintenance. Our development timeline, we launched in 2021 in 2022 built some of the prototypes that you see, saw in the videos, developed some of the hardware. We are closing our seed round now in early 2024 and that'll get our CT system on the market, our Mr system closer to a design freeze. And ultimately, we want to reach sales of our Mr system by 2026 initially, we can leverage existing coding, and this is a really key point. So Mr liver biopsies, hospital reimbursement, around 20k Mr liver ablation around 50k the same time, many facilities are not doing these procedures because they're quite challenging. As I described, you're moving the patient in and out. It requires a very experienced physician. It takes a long time. So they're using their scanners for diagnostic MRI revenue, there is about two to 5k with our system, by simplifying standardizing the procedure, we can enable these facilities to do these much higher revenue generating procedures, bring them to the in Rens MRI, increasing their MRI revenue about 10x we've got a great team. We're a spin out from Stanford University at Northeastern we've got industry experts, engineers, doctors, and we're supported and funded by the groups you see on the screen. Thank you. Applause.