ICU LLM: Reducing Alarm Fatigue
Advanced Data Analytics Group Project: Omar Aamir, Sarika Gaind, Sana Ambreen
The Problem
Current ICU monitoring systems track vitals like heart rate and oxygen levels using static thresholds. This often leads to "false alarms" and clinician burnout. Our project implements a Large Language Model (LLM) to provide context-aware monitoring, filtering out noise and focusing on true patient distress.
Project Data Access
Below are links to the research datasets. Each folder contains its own index of PDF files.