Mantis Biotech is making ‘digital twins’ of humans to help solve medicine’s data availability problem

Mantis Biotech, a New York-based company, has developed a platform that creates synthetic datasets to help solve the data availability problem in medicine. This issue arises when large language models struggle to process edge cases, such as rare diseases and unusual conditions, where reliable data is scarce. The platform integrates disparate sources of data to build so-called "digital twins" of the human body: physics-based, predictive models of anatomy, physiology, and behavior. These digital twins can be used for various medical applications, including studying and testing new procedures, training surgical robots, and simulating medical issues.

The platform takes data from various sources, including textbooks, motion capture cameras, biometric sensors, training logs, and medical imaging. It then uses an LLM-based system to route, validate, and synthesize the data streams, running them through a physics engine to create high-fidelity renders of the dataset. This can be used to train predictive models that can make accurate predictions about human performance.

Mantis' technology has the potential to revolutionize biomedical research by providing more accurate and reliable data. For instance, it can be used to predict the likelihood of an athlete developing an injury based on their recent performance, training load, diet, and other factors. The startup recently raised $7.4 million in seed funding to further develop its platform.

The company's platform enhances available information by grounding the generated synthetic data in realistic physics models of anatomy. This allows for more accurate predictions and simulations, which can be used to improve medical research and treatment.