Saved in:
Bibliographic Details
Main Authors: Bakumenko, Alexander, Hoelscher, Janine, Smith, Hudson
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.15847
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917093853626368
author Bakumenko, Alexander
Hoelscher, Janine
Smith, Hudson
author_facet Bakumenko, Alexander
Hoelscher, Janine
Smith, Hudson
contents Early identification of intensive care patients at risk of in-hospital mortality enables timely intervention and efficient resource allocation. Despite high predictive performance, existing machine learning approaches lack transparency and robustness, limiting clinical adoption. We present a lightweight, transparent multimodal ensemble that fuses physiological time-series measurements with unstructured clinical notes from the first 48 hours of an ICU stay. A logistic regression model combines predictions from two modality-specific models: a bidirectional LSTM for vitals and a finetuned ClinicalModernBERT transformer for notes. This traceable architecture allows for multilevel interpretability: feature attributions within each modality and direct per-case modality attributions quantifying how vitals and notes influence each decision. On the MIMIC-III benchmark, our late-fusion ensemble improves discrimination over the best single model (AUPRC 0.565 vs. 0.526; AUROC 0.891 vs. 0.876) while maintaining well-calibrated predictions. The system remains robust through a calibrated fallback when a modality is missing. These results demonstrate competitive performance with reliable, auditable risk estimates and transparent, predictable operation, which together are crucial for clinical use.
format Preprint
id arxiv_https___arxiv_org_abs_2511_15847
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Transparent Early ICU Mortality Prediction with Clinical Transformer and Per-Case Modality Attribution
Bakumenko, Alexander
Hoelscher, Janine
Smith, Hudson
Machine Learning
Early identification of intensive care patients at risk of in-hospital mortality enables timely intervention and efficient resource allocation. Despite high predictive performance, existing machine learning approaches lack transparency and robustness, limiting clinical adoption. We present a lightweight, transparent multimodal ensemble that fuses physiological time-series measurements with unstructured clinical notes from the first 48 hours of an ICU stay. A logistic regression model combines predictions from two modality-specific models: a bidirectional LSTM for vitals and a finetuned ClinicalModernBERT transformer for notes. This traceable architecture allows for multilevel interpretability: feature attributions within each modality and direct per-case modality attributions quantifying how vitals and notes influence each decision. On the MIMIC-III benchmark, our late-fusion ensemble improves discrimination over the best single model (AUPRC 0.565 vs. 0.526; AUROC 0.891 vs. 0.876) while maintaining well-calibrated predictions. The system remains robust through a calibrated fallback when a modality is missing. These results demonstrate competitive performance with reliable, auditable risk estimates and transparent, predictable operation, which together are crucial for clinical use.
title Transparent Early ICU Mortality Prediction with Clinical Transformer and Per-Case Modality Attribution
topic Machine Learning
url https://arxiv.org/abs/2511.15847