This is the third installment in a six-part series examining the impact of AI on medical research and treatments. Audra Moran, head of the Ovarian Cancer Research Alliance, emphasizes that ovarian cancer is "rare, underfunded, and deadly. " Early detection is crucial, but ovarian cancer usually starts in the fallopian tubes, often spreading before symptoms arise. AI-powered blood tests are emerging to detect such cancers in their early stages, potentially saving lives. Dr. Daniel Heller from Memorial Sloan Kettering Cancer Center is developing technology using carbon nanotubes to identify cancer-specific patterns in blood samples. While these patterns are too subtle for humans to discern, AI algorithms can interpret them by learning from data. The challenge is the rarity of ovarian cancer, leading to limited data for AI training. Despite this, initial AI efforts have surpassed existing cancer biomarkers in accuracy. Heller aims to create a tool for diagnosing gynecological diseases more rapidly, which might be feasible in three to five years. AI is also streamlining other blood tests.

For instance, Karius in California uses AI to quickly identify pneumonia pathogens, significantly reducing testing costs and time. Their approach, which compares patient samples with a vast microbial DNA database, is only possible due to AI. However, understanding AI’s connections between biomarkers and diseases remains a mystery. Dr. Slavé Petrovski's AI platform Milton identifies diseases using biomarkers with high success, demonstrating AI's ability to discern complex patterns. Sharing and accessing data remains a challenge, yet organizations like Ocra are working to create patient registries to enhance AI training. As Moran notes, AI in medical research is still in its early, evolving stages.