How SHEPHERD AI Can Help with Quicker Rare Disease Diagnoses

March Forward

Send us a text

The search for a rare disease diagnosis can be a long and lonely journey. But what if advanced technology could help light the way? 🔍

In this episode of March Forward, we're exploring SHEPHERD, an innovative AI approach bringing new hope to rare disease diagnosis. SHEPHERD uses a clever "few-shot learning" method, allowing it to learn effectively even from limited patient information. It also uses a deep rare disease knowledge graph, acting like a vast, interconnected map to find important connections for your unique health picture. 🧠

This isn't just theory. SHEPHERD has been tested on real-world cases from the Undiagnosed Diseases Network, successfully identifying causal genes in 40% of patients across many challenging areas. 🧬 It works with clinicians, providing vital insights to shorten the diagnostic odyssey. ✨

Learn More About SHEPHERD's Research:

• SHEPHERD's Published Research (npj Digital Medicine): https://www.nature.com/articles/s41746-025-01749-1
• Zitnik Lab SHEPHERD Project Page: https://zitniklab.hms.harvard.edu/projects/SHEPHERD/
• SHEPHERD Preprint (medRxiv): https://www.medrxiv.org/content/10.1101/2022.12.07.22283238v2
• PubMed Central Article: https://pmc.ncbi.nlm.nih.gov/articles/PMC12222559/

Subscribe to March Forward, for more deep dives. Like this episode, and share your thoughts below; we’re here to listen.

#RareDisease #SHEPHERD #AIHealthcare #DiagnosticOdyssey #RareDiseaseAwareness #GeneticDiagnosis #RareDiseaseResearch

March Forward - Podcast by March.ai