Tech • IA • Crypto
Pourquoi 80% des essais cliniques foirent ?
8/10TL;DR
A 24-year-old founder has raised $7.4 million to build high-fidelity digital human clones aimed at transforming clinical trials and personalized medicine.
KEY POINTS
Massive inefficiency in clinical trials
Around 80% of clinical trials face delays due to poor-quality or insufficient data, with each failed trial costing roughly $15 million. This leads to promising treatments being abandoned and patients waiting for therapies that may never reach the market.
Georgia Witchell and Mentis Biotech
Georgia Witchell, founder of Mentis Biotech, has secured $7.4 million in funding to tackle this systemic issue. Her company focuses on creating detailed digital replicas of human bodies to improve medical testing and prediction.
Digital twin technology
The platform aggregates diverse datasets including medical imaging, biometric sensors, and motion capture. An AI system validates and integrates the data, while a physics engine builds a high-fidelity “digital twin” that mimics real biological responses.
Predictive capabilities already in use
The technology is already being tested in elite sports, including by an NBA team. It can reportedly predict injuries such as Achilles tendon ruptures weeks in advance by analyzing movement patterns, recovery, and sleep data.
Applications in medicine
The broader goal is medical transformation: simulating drug responses before patient use, rehearsing surgeries on virtual replicas, and training robotic surgeons on synthetic cases, including rare diseases where data is scarce.
Addressing the data gap in AI healthcare
The core problem is not scientific knowledge but a lack of high-quality training data for medical AI. Digital twins could generate scalable, realistic datasets to accelerate innovation and reduce risk.
CONCLUSION
By combining AI with physics-based simulation, Mentis Biotech aims to reduce clinical trial failures and enable safer, more personalized healthcare at scale.