Advancing Rare Disease Diagnosis Through AI and Genomic Innovation

This episode of the Becker's Healthcare Podcast explores the transformative potential of AI and genomic innovation in advancing rare disease diagnosis, particularly in pediatric care. Dr. Shannon Heyman from Lurie Children's Hospital and Dr. Calum Yakubian from IQVIA discuss the harrowing 'diagnostic odyssey' faced by families of children with rare genetic diseases—where 40% remain undiagnosed and the average diagnostic delay is five years. They highlight how whole genome sequencing, when used as a first-line test, can dramatically increase diagnostic yield, especially when supported by multidisciplinary teams and AI-driven tools. The conversation centers on how natural language processing (NLP) and AI are being used to extract meaningful phenotypic data from unstructured medical records, reducing the signal-to-noise problem in genomic data and accelerating diagnosis. The panel emphasizes that successful implementation hinges not just on technology, but on clinical buy-in, governance, human-in-the-loop design, and embedding AI into existing workflows. They caution against 'bright shiny object' syndrome, stressing that AI must solve well-defined clinical problems with measurable impact. Looking ahead, they envision AI-enabled newborn screening at scale, making early diagnosis routine and accessible to thousands of families.