HIST-AID: Leveraging Historical Patient Reports for Enhanced Multi-Modal Automatic Diagnosis

TL;DR

HIST-AID introduces a framework that leverages five years of patient historical reports and scans to significantly improve the accuracy of AI-based chest X-ray diagnosis, demonstrating consistent gains across diverse patient groups.

Contribution

This work presents HIST-AID, a novel framework that incorporates historical patient data into AI diagnosis models, enhancing accuracy beyond current scan-only approaches.

Findings

AUROC increased by 6.56% with HIST-AID

AUPRC increased by 9.51% with HIST-AID

Historical data improves diagnostic reliability across demographics

Abstract

Chest X-ray imaging is a widely accessible and non-invasive diagnostic tool for detecting thoracic abnormalities. While numerous AI models assist radiologists in interpreting these images, most overlook patients' historical data. To bridge this gap, we introduce Temporal MIMIC dataset, which integrates five years of patient history, including radiographic scans and reports from MIMIC-CXR and MIMIC-IV, encompassing 12,221 patients and thirteen pathologies. Building on this, we present HIST-AID, a framework that enhances automatic diagnostic accuracy using historical reports. HIST-AID emulates the radiologist's comprehensive approach, leveraging historical data to improve diagnostic accuracy. Our experiments demonstrate significant improvements, with AUROC increasing by 6.56% and AUPRC by 9.51% compared to models that rely solely on radiographic scans. These gains were consistently observed across diverse demographic groups, including variations in gender, age, and racial categories. We show that while recent data boost performance, older data may reduce accuracy due to changes in patient conditions. Our work paves the potential of incorporating historical data for more reliable automatic diagnosis, providing critical support for clinical decision-making.