Shyam Natarajan, Avenda Health - 3D AI Mapping for Cancer | LSI USA '24

Shyam Natarajan  0:03  
I'm here as co founder and CEO of Avenda health. We are a commercial stage medtech company using multimodal, spatial AI to untangle the complexities of cancer care. So I started my career actually as an academic professor of bioengineering Urology at UCLA. We spun out of UCLA and went through the med tech innovator program. So first time founder, but my career was really spent in understanding how to better diagnose and manage and treat prostate cancer, really focusing on what is the big challenge of imaging. You know, treatment planning, decision support. And so through that experience and over 10 years of NIH funded research, we've built the largest data set of its kind using multiple kinds of information, imaging, biopsy, pathology, biomarkers, et cetera, and have created the first multimodal spatial AI platform to power the future of cancer care. We are FDA cleared, have breakthrough device designation, and have applied and been issued a new CPD code, the first AI code for this space. Our goal is to get to build a foundational cancer AI model to go beyond prostate cancer and tackle the 100 billion dollar cancer market, but really starting with prostate cancer first, because it is the most common problem in the cancer space. One out of every seven eight men. Prostate cancer is a unique kind of cancer in that imaging and biopsy and diagnostics aren't reliable alone in actually determining the extent of disease. About 68% of the disease by volume is actually invisible to imaging. This is very impactful for patients, because ends up one out of five or one out of two end up having residual cancer post treatment. And in fact, many of these patients end up losing their quality of life because of that lack of spatial identification of where the tumor actually is both within and around the prostate, so the standard of care is quite poor. Here you there's no one standard patient can either go on surveillance, they can get ablative therapies, or they can get surgery and radiation, regardless of the treatment modality, patients still suffer quality of life complications, but importantly, cancer is still left behind because of that lack of visualization. So meet unfold. Ai, a new superpower for oncologist. This is the first multimodal spatial AI platform that is cloud based, that enables a urologist, who doesn't normally read images, to synthesize and aggregate all of the patient's available diagnostic information to make a better determination and to actually guide their therapies to deliver better outcomes. So here's here's what it looks like compared to the standard of the care. MRI only detects a hot spot and requires an expert radiologist to contour and very precisely understand where the cancer is however, because it underestimates the true extent of disease and because there's a lot of rich information available in the patient's diagnostic history unfold. Ai brings that information together, creates a 3d cancer probability map showing the extent of disease as well as the optimal margin. We validated this with surgical prostectomy specimens as well as biopsy specimens. On a large data set of hundreds of 1000s of data points today, we have over 4 million. So how it works is that a patient gets their existing diagnostic tests and information they don't need any new diagnosis. We use that information and sync not only with healthcare enterprise in the IT space, but also we're able to derive metadata from unconnected edge devices and spatially map all of that information together. This prostate cancer probability map allows us to actually create a whole new modality, a synthetic imaging modality, that has led to numerous AI biomarkers that have shown superiority to the standard of care. So how do you use this information? Well, doctors today are using this for patient counseling, decision support, and actually walking through a diagnosis of prostate cancer and understanding, is it surgery? Is it radiation? What is the best treatment for them? They're using this to guide ablative therapies and actually overlay that information as they're ablating to make sure that they're not leaving any cancer behind. They're also using this for surgical guidance. We can actually bring this into the robot and use this as a real time navigation aid in order to make sure that when a doctor spares nerves, they're not leaving any cancer behind. So this is actionable and clinically meaningful. We've shown an improvement over the standard of care in terms of sensitivity. This was 10 physicians doing 1000 reads from 38% sensitivity to 97% sensitivity. We. They also changed the recommendation on how to deliver care for the patient in about a third of cases, and actually over half the time, they were still making changes, but deciding whether to treat more or to treat less or to make subtle decision aids in their current treatment category. We also have shown emergent abilities and new biomarkers from this rich data set. We didn't train the model originally to detect cancer outside of the prostate, but it turns out it can. So this is detecting extra capsular disease, which is the most common challenge in surgery today. In fact, this is clinically meaningful, because this is a huge problem in the healthcare system. Salvage therapy costs the healthcare $25,000 per patient, and we've shown in studies that we can reduce the incidence by half. Now we are the first to market. This is a new category, similar to genomic biomarkers and advanced imaging. So it's priced as such as a per patient analysis, or per click. We're targeting urology clinics first, as well as hospitals, and we're compatible with most of the systems on the market, so we're platform agnostic. Regardless of what imaging provider they got. We also have a clear path to reimbursement. We just received a new CPT code supported by seven different societies, including radiology urology and radiation oncology, three societies that never agree on anything. We also got a augmentative AI code, which is a higher bar. We also submitted a new tech, APC, and we expect a hospital outpatient payment in July for around $5,000 or so per analysis and coverage will be supported physician office later, and coverage will be supported by our robust clinical evidence package as well as ongoing studies. So since we soft launch last year, we've analyzed about 1000 patients. We have active customer contracts. Our first customer was actually UCLA health, and they just renewed. We've been able to not only gather a lot of data points and develop a lot of biomarkers, we also have significant inbound interests, about 200 patients on a wait list representing representing a huge, huge market, but on the subset that we've started contracting with represents a pipeline of over 10 million in annual revenue. Our clinical studies were also supported by the NIH as well as large grants. So our vision for the future really is to use this multimodal spatial AI capability to not just improve prostate cancer, but also many other kinds of solid organ tumors. And so rather like chatgpt Being a foundational model for text, we're building the largest foundational model for solid organ tumors in cancer care, we've been supported by a strong advisory board as well as strong board members as well. And to date, we've raised about $16 million in equity, another $3 million in non dilutive and we're out here kicking off our Series C, raising $30 million for commercialization as well as going into new indications today. So event to health really is about bringing better care to patients and helping physicians go beyond their abilities. So superhuman abilities for cancer care. Thank you. Applause.