IMMUNOLOGY2026™ Conference Recordings For Attendees-Machine learning to advance rheumatoid arthritis treatment strategies

Video Transcription

Video Summary

The speaker discusses rheumatoid arthritis (RA), a common autoimmune joint disease, and the challenge of choosing the best treatment for each patient. Although there are many therapies, clinicians still rely on trial and error. The talk explains how machine learning and deep phenotyping can help identify predictors of treatment response by analyzing electronic health records, clinical features, and synovial tissue cell profiles. A key issue is the “curse of dimensionality”: there are many variables but relatively few patients with measured outcomes. To address this, the speaker describes clustering synovial cell data into six abundance phenotypes and linking them to treatment patterns. The talk also shows how EHR-based algorithms can identify large RA cohorts and emulate clinical trials, despite missing standardized disease activity data. The overall goal is to connect biological and clinical data to better match patients with the right RA therapy.

Meta Tag

Date April 16, 2026 8:45 AM - 9:07 AM

Room 205

Session Artificial Intelligence Approaches in Immune-Mediated Disease, Sponsored by DAIT, AMIB, NIAID, NIH

Speaker Katherine Liao

Track Computational and Systems Immunology (COMP)

Year 2026

Keywords

rheumatoid arthritis

machine learning

treatment response

deep phenotyping

electronic health records

April 16, 2026 8:45 AM - 9:07 AM

205

Artificial Intelligence Approaches in Immune-Mediated Disease, Sponsored by DAIT, AMIB, NIAID, NIH

Katherine Liao

Computational and Systems Immunology (COMP)

2026