Early Prediction of Sepsis: Feature-Aligned Transfer Learning

TL;DR

This work tackles the urgent need for early sepsis prediction by proposing Feature-Aligned Transfer Learning (FATL), a framework that aligns consistently reported features across diverse studies and combines knowledge from multiple source models through feature-focused weight transfer. FATL aims to improve generalization across populations and clinical settings, addressing heterogeneity in data and reducing alarm fatigue by emphasizing robust, reproducible signals. Preliminary findings identify a core subset of overlapping features spanning vital signs, labs, and demographics, suggesting a practical path for resource-limited environments. The approach holds promise for faster, more equitable early detection of sepsis with potential to reduce mortality and healthcare costs, contingent on robust external validation and careful integration into clinical workflows.

Abstract

Sepsis is a life threatening medical condition that occurs when the body has an extreme response to infection, leading to widespread inflammation, organ failure, and potentially death. Because sepsis can worsen rapidly, early detection is critical to saving lives. However, current diagnostic methods often identify sepsis only after significant damage has already occurred. Our project aims to address this challenge by developing a machine learning based system to predict sepsis in its early stages, giving healthcare providers more time to intervene. A major problem with existing models is the wide variability in the patient information or features they use, such as heart rate, temperature, and lab results. This inconsistency makes models difficult to compare and limits their ability to work across different hospitals and settings. To solve this, we propose a method called Feature Aligned Transfer Learning (FATL), which identifies and focuses on the most important and commonly reported features across multiple studies, ensuring the model remains consistent and clinically relevant. Most existing models are trained on narrow patient groups, leading to population bias. FATL addresses this by combining knowledge from models trained on diverse populations, using a weighted approach that reflects each models contribution. This makes the system more generalizable and effective across different patient demographics and clinical environments. FATL offers a practical and scalable solution for early sepsis detection, particularly in hospitals with limited resources, and has the potential to improve patient outcomes, reduce healthcare costs, and support more equitable healthcare delivery.

Figures (5)

• Figure 1: Introduction of sepsis
• Figure 2: Comparison of mortality and early prediction for patients
• Figure 3: The architecture of our system
• Figure 4: Conceptual Architecture Diagram of FATL
• Figure 5: Overlapping Features