RadFusion: Benchmarking Performance and Fairness for Multimodal Pulmonary Embolism Detection from CT and EHR

TL;DR

RadFusion introduces a comprehensive multimodal dataset combining CT scans and EHR data for pulmonary embolism detection, enabling evaluation of model performance and fairness across diverse patient groups.

Contribution

This work provides the RadFusion benchmark dataset and evaluates multimodal fusion models for pulmonary embolism detection, addressing the gap in fairness and clinical context integration.

Findings

Multimodal models improve detection accuracy.

Integrating EHR and imaging data enhances robustness.

Fairness across demographic groups is maintained.

Abstract

Despite the routine use of electronic health record (EHR) data by radiologists to contextualize clinical history and inform image interpretation, the majority of deep learning architectures for medical imaging are unimodal, i.e., they only learn features from pixel-level information. Recent research revealing how race can be recovered from pixel data alone highlights the potential for serious biases in models which fail to account for demographics and other key patient attributes. Yet the lack of imaging datasets which capture clinical context, inclusive of demographics and longitudinal medical history, has left multimodal medical imaging underexplored. To better assess these challenges, we present RadFusion, a multimodal, benchmark dataset of 1794 patients with corresponding EHR data and high-resolution computed tomography (CT) scans labeled for pulmonary embolism. We evaluate several representative multimodal fusion models and benchmark their fairness properties across protected subgroups, e.g., gender, race/ethnicity, age. Our results suggest that integrating imaging and EHR data can improve classification performance and robustness without introducing large disparities in the true positive rate between population groups.