Case Study: Moça achieves scalable early cognitive decline detection with OSEDEA’s AI-powered MoCA Solo

Moça achieves over 90% AI scoring accuracy with OSEDEA

The customer, Moça, the creator of the Montreal Cognitive Assessment (MoCA) test, faced a challenge in scaling their gold-standard cognitive evaluation. They sought to develop a fully autonomous digital application, MoCA Solo, but needed to accurately and automatically score patient audio and visual responses without human annotators, a task requiring strict clinical precision. OSEDEA was engaged to engineer the AI models for this purpose.

OSEDEA implemented a solution using AI models for speech-to-text and visual recognition, integrating technologies like Whisper and YOLO. They achieved over 90% accuracy in scoring with just 200 training samples. The solution provided Moça with a clinically reliable, scalable application and comprehensive documentation for future regulatory approval. OSEDEA's work enabled early, consistent cognitive decline screening, making the test more accessible.

Moça