Using Artificial Intelligence for Rare Dermatologic Diseases: Getting the Right Patients the Right Treatment at the Right Time

The presentation discusses the use of artificial intelligence (AI) to improve the diagnosis and treatment of rare dermatologic diseases. The speaker, not a dermatologist but experienced in AI and data, highlights the collaboration between their organization and dermatologists to analyze extensive patient data from a registry called DataDerm, which contains millions of clinical encounters. They focus on developing a tool called Patient Finder, which utilizes AI to identify patterns within patient data, helping find undiagnosed or misdiagnosed patients by uncovering commonalities in their histories. The AI tool, known as Phenom, performs digital phenotyping to detect these patterns, akin to identifying fingerprints in complex datasets. By leveraging unstructured clinical notes, the technology enables the extraction of insightful data, such as disease activity estimates, allowing researchers to amplify the endpoints available for studying various conditions. The speaker emphasizes that this method could accelerate diagnosis and treatment and improve the understanding of disease progression. They also stress the importance of explainability in AI, allowing clinicians to understand why certain patients are flagged for further evaluation. Ultimately, this initiative aims to enhance patient care by ensuring the right patients receive the right treatments at the right time.