How AI Helps Solve Medical Mysteries at Boston Children’s Hospital | OpenAI Forum

This panel features a discussion on a research project between OpenAI, Harvard, and the Manton Center for Orphan Disease Research at Boston Children's Hospital. The panel highlights the challenge of diagnosing rare diseases, where the average diagnostic journey for a patient can span 6–7 years. The participants discuss how AI models were used to analyze complex genetic data, identifying potential leads for rare disease diagnoses by rapidly cross-referencing patient data with medical literature and databases.

The discussion covers the practical workflow, starting with the patient experience and moving into the technical implementation of large language models (LLMs). The speakers detail how they transitioned from analyzing known cases to tackling unsolved ones, emphasizing that the primary value of AI is not in replacing the human diagnostician, but in acting as a powerful tool to process massive datasets and prioritize the most promising leads. They also touch upon future directions, including the broader application of AI in medicine and the importance of making such diagnostic tools more accessible to a wider patient and medical population.