Early Detection of Pancreatic Cancer Using Multimodal Learning on Electronic Health Records

TL;DR

This study introduces a multimodal machine learning approach that combines diagnosis history and lab data from electronic health records to detect pancreatic cancer early, significantly outperforming existing methods and identifying relevant biomarkers.

Contribution

The paper presents a novel multimodal deep learning framework integrating neural controlled differential equations and cross-attention for early PDAC detection from EHR data.

Findings

Achieved 6.5% to 15.5% higher AUC than previous methods.

Identified both known and new biomarkers associated with PDAC risk.

Demonstrated effectiveness on a real-world dataset of nearly 4,700 patients.

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest cancers, and early detection remains a major clinical challenge due to the absence of specific symptoms and reliable biomarkers. In this work, we propose a new multimodal approach that integrates longitudinal diagnosis code histories and routinely collected laboratory measurements from electronic health records to detect PDAC up to one year prior to clinical diagnosis. Our method combines neural controlled differential equations to model irregular lab time series, pretrained language models and recurrent networks to learn diagnosis code trajectory representations, and cross-attention mechanisms to capture interactions between the two modalities. We develop and evaluate our approach on a real-world dataset of nearly 4,700 patients and achieve significant improvements in AUC ranging from 6.5% to 15.5% over state-of-the-art methods. Furthermore, our model identifies diagnosis codes and laboratory panels associated with elevated PDAC risk, including both established and new biomarkers. Our code is available at https://github.com/MosbahAouad/EarlyPDAC-MML.