From Discharge Letters to Process Traces with LLMs: A Human-in-the-Loop Pipeline...36th Medical Informatics Europe (MIE) Conference, May 25-28, 2026, Genoa, Italy.

We present a modular human-in-the-loop pipeline that converts unstructured discharge letters into process-mining–ready event traces. The pipeline uses task-specific Large Language Model prompting to extract temporally ordered clinical events, followed by expert validation and iterative refinement. V...

Descripción completa

Detalles Bibliográficos
Publicado en:Studies in Health Technology & Informatics Vol. 336; pp. 1037 - 1039
Autores principales: BOTTRIGHI, Alessio, CANESSA, Alessandro, FERRANDI, Delfina, LEONARDI, Giorgio, MACONI, Antonio, MASSARINO, Costanza, MONTANI, Stefania, ROVETA, Annalisa, STRIANI, Manuel
Formato: pictorial proceedings tables/charts Journal Article
Publicado: Sage Publications Inc. 2026
Acceso en línea:Ver este registro en EBSCOhost
Descripción
Sumario:We present a modular human-in-the-loop pipeline that converts unstructured discharge letters into process-mining–ready event traces. The pipeline uses task-specific Large Language Model prompting to extract temporally ordered clinical events, followed by expert validation and iterative refinement. Validated traces are stored for PM analyses. The approach is currently evaluated on 466 Stroke Unit discharge letters from Alessandria Hospital.