ISeeU2: Visually Interpretable ICU mortality prediction using deep learning and free-text medical notes

TL;DR

This paper presents ISeeU2, a deep learning model that predicts ICU mortality from medical notes with high accuracy and provides visual explanations to improve interpretability and clinical trust.

Contribution

The study introduces a novel deep learning approach that uses raw nursing notes for mortality prediction and offers visual interpretability tools, outperforming traditional scoring methods.

Findings

Achieved ROC of 0.8629, surpassing SAPS-II score.

Provided visual explanations for model predictions.

Enhanced interpretability of deep learning in clinical settings.

Abstract

Accurate mortality prediction allows Intensive Care Units (ICUs) to adequately benchmark clinical practice and identify patients with unexpected outcomes. Traditionally, simple statistical models have been used to assess patient death risk, many times with sub-optimal performance. On the other hand deep learning holds promise to positively impact clinical practice by leveraging medical data to assist diagnosis and prediction, including mortality prediction. However, as the question of whether powerful Deep Learning models attend correlations backed by sound medical knowledge when generating predictions remains open, additional interpretability tools are needed to foster trust and encourage the use of AI by clinicians. In this work we show a Deep Learning model trained on MIMIC-III to predict mortality using raw nursing notes, together with visual explanations for word importance. Our model reaches a ROC of 0.8629 (+/-0.0058), outperforming the traditional SAPS-II score and providing enhanced interpretability when compared with similar Deep Learning approaches.