Development and Validation of an LLM-Based Pipeline for Transcranial Doppler Interpretation in Pediatric Sickle Cell Disease​

Abstract

Slide deck presented by Sahal Master, BDS, MPH at the American Medical Informatics Association (AMIA) 2026 Clinical Informatics Conference. This presentation describes the work done at the PEDSnet learning health system to develop an LLM pipeline for the extraction of measurements from clinical notes data as part of a study with PEDSnet collaborator, Dr. Aleksandra Dain. Specifically, this case study evaluated an LLM pipeline for artery-level velocity extraction for transcranial doppler outcome classification. Funding Statement The research reported in this presentation was conducted using PEDSnet, A Pediatric Clinical Research Network. PEDSnet has been developed with funding from the Patient-Centered Outcomes Research Institute (PCORI); PEDSnet’s participation in PCORnet is funded through PCORI award RI-CHOP-01-PS10. This presentation includes data from the Children’s Hospital of Philadelphia. Research reported in this presentation was supported by grant 1P30HS029755 funded by the Agency of Healthcare Research and Quality (AHRQ) and the Patient-Centered Outcomes Research Institute (PCORI®) through a research collaboration. The opinions presented in this presentation are solely the responsibility of the author(s) and do not necessarily represent the views of AHRQ, the U.S. Department of Health and Human Services, or PCORI®.