Creative genius. No matter how you define it, it usually requires at least three ingredients, according to Walter Isaacson’s book on Leonardo da Vinci: “passionate curiosity, careful observation, and an imagination so playful that it flirts with fantasy.” DaVinci had them. So did Albert Einstein and Steve Jobs. Successful startups, including those in health care, often have a combination of those traits. In fact, these traits often separate winners from losers.
With such aspirations in mind, Mayo Clinic Platform recently launched Mayo Clinic Platform Accelerate, the goal of which is to help innovative startups get market-ready. The 20-week program recently chose its first four candidates: Seer Medical, Quadrant Health, ScienceIO and cliexa. Each will receive a benefits package that includes access to Mayo Clinic’s de-identified data sets, validation frameworks, clinical workflow planning, and mentorship. Mayo Clinic in turn will take an equity position in each startup.
Seer is reimagining how people undergo long-term monitoring of epilepsy, sleep, and cardiac conditions for faster paths to diagnoses and better patient outcomes. The first international company to join Mayo Clinic Platform Accelerate, Seer is working to innovate and strengthen its models for epilepsy monitoring and prediction. Based in Melbourne, Australia, the company is moving patient monitoring from hospital to home.
Suggesting that epileptic seizures might be predicted before they happen with home-based monitoring might seem like a flight of fantasy, but preliminary research suggests it’s doable. In the past, the only way to forecast the onset of a seizure was by capturing EEG data through an implantable intracranial device. But the latest studies suggest that a novel sub-scalp monitoring system may be a plausible alternative to predict focal seizure activity.
Quadrant Health is testing its predictive models for various conditions, including sepsis, Clostridioides difficile, falls, and readmissions, with the help of Mayo Clinic’s robust data sets. Quadrant’s founding team members are all current or former Stanford students and have been working on this project for the past two years. Their AI platform uses 4,300+ clinical and demographic features per analysis to make its predictions, and generates a risk score that clinicians and patients can comprehend without an understanding of data science — what they are calling explainable AI.
ScienceIO will harness the power of Mayo Clinic Platform to leverage Mayo Clinic’s data sets and derive new insights from unstructured patient data. The startup’s technology converts unstructured data into contextualized information and is capable of detecting and extracting more than 9 million health care concepts, clinical variables, and medical codes in text. This enables its digital tools to capture a broad spectrum of information, spanning medical conditions, devices, procedures, vitals, biomarkers, mutations, and more, according to the company’s website.
cliexa works to remove barriers that have inhibited providers from unleashing the power of patient-reported data; its goal is to improve clinical care and outcomes. During the 20-week program, cliexa is working on expanding its opioid prediction model as well as new models for the prediction of cardiovascular risk. cliexa describes itself as a provider of integrated care solutions, with the ability to collect and analyze patient-reported data to generate actionable insights. The startup offers AI-enabled patient triaging and onboarding, integration with a health care system’s EHR, quality and compliance tools, and remote patient monitoring.
While many health care startups begin as daydreams in the minds of creative geniuses, entrepreneurs, and technologists with the right blend of passion, observational skills, and imagination are the future of digital medicine.
Written by John Halamka, MD, president, and Paul Cerrato, senior research analyst and communications specialist at Mayo Clinical Platform, this piece was originally posted to their blog page, Digital Health Frontier.