Mantis Biotech is developing a platform that uses artificial intelligence to create “digital twins” of the human body, aiming to address data limitations in biomedical research. The approach combines large language models with physics-based simulations to generate synthetic datasets for healthcare applications.
The system aggregates data from multiple sources, including medical imaging, biometric sensors, motion capture, and clinical records. It then processes and validates the information using AI models before applying a physics engine to generate realistic simulations of human anatomy and behavior. These digital representations can be used to train predictive models and analyze complex biological scenarios.
The technology is designed to overcome gaps in traditional datasets, particularly in edge cases such as rare diseases or uncommon conditions where data is scarce. By generating synthetic data, the platform enables researchers to simulate scenarios that would otherwise be difficult or ethically constrained to study.
Potential applications include drug discovery, clinical decision support, surgical training, and real-time diagnostics. The platform is also being used in professional sports to model athlete performance and predict injury risks based on physical and behavioral data.
Mantis has raised $7.4 million in seed funding from investors including Decibel VC and Y Combinator. The company plans to expand its platform for broader use in preventative healthcare and pharmaceutical research, including support for clinical trials.
The development reflects growing interest in combining AI with simulation-based modeling to enhance biomedical research, as companies seek new ways to generate high-quality data and accelerate innovation in healthcare.
