Causal Debiasing Medical Multimodal Representation Learning with Missing Modalities

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhu, Xiaoguang, Sun, Lianlong, Liu, Yang, Jiang, Pengyi, Srivatsa, Uma, Chiamvimonvat, Nipavan, Filkov, Vladimir
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915483030126592
author Zhu, Xiaoguang
Sun, Lianlong
Liu, Yang
Jiang, Pengyi
Srivatsa, Uma
Chiamvimonvat, Nipavan
Filkov, Vladimir
author_facet Zhu, Xiaoguang
Sun, Lianlong
Liu, Yang
Jiang, Pengyi
Srivatsa, Uma
Chiamvimonvat, Nipavan
Filkov, Vladimir
contents Medical multimodal representation learning aims to integrate heterogeneous clinical data into unified patient representations to support predictive modeling, which remains an essential yet challenging task in the medical data mining community. However, real-world medical datasets often suffer from missing modalities due to cost, protocol, or patient-specific constraints. Existing methods primarily address this issue by learning from the available observations in either the raw data space or feature space, but typically neglect the underlying bias introduced by the data acquisition process itself. In this work, we identify two types of biases that hinder model generalization: missingness bias, which results from non-random patterns in modality availability, and distribution bias, which arises from latent confounders that influence both observed features and outcomes. To address these challenges, we perform a structural causal analysis of the data-generating process and propose a unified framework that is compatible with existing direct prediction-based multimodal learning methods. Our method consists of two key components: (1) a missingness deconfounding module that approximates causal intervention based on backdoor adjustment and (2) a dual-branch neural network that explicitly disentangles causal features from spurious correlations. We evaluated our method in real-world public and in-hospital datasets, demonstrating its effectiveness and causal insights.
format Preprint
id arxiv_https___arxiv_org_abs_2509_05615
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Causal Debiasing Medical Multimodal Representation Learning with Missing Modalities
Zhu, Xiaoguang
Sun, Lianlong
Liu, Yang
Jiang, Pengyi
Srivatsa, Uma
Chiamvimonvat, Nipavan
Filkov, Vladimir
Machine Learning
Artificial Intelligence
Medical multimodal representation learning aims to integrate heterogeneous clinical data into unified patient representations to support predictive modeling, which remains an essential yet challenging task in the medical data mining community. However, real-world medical datasets often suffer from missing modalities due to cost, protocol, or patient-specific constraints. Existing methods primarily address this issue by learning from the available observations in either the raw data space or feature space, but typically neglect the underlying bias introduced by the data acquisition process itself. In this work, we identify two types of biases that hinder model generalization: missingness bias, which results from non-random patterns in modality availability, and distribution bias, which arises from latent confounders that influence both observed features and outcomes. To address these challenges, we perform a structural causal analysis of the data-generating process and propose a unified framework that is compatible with existing direct prediction-based multimodal learning methods. Our method consists of two key components: (1) a missingness deconfounding module that approximates causal intervention based on backdoor adjustment and (2) a dual-branch neural network that explicitly disentangles causal features from spurious correlations. We evaluated our method in real-world public and in-hospital datasets, demonstrating its effectiveness and causal insights.
title Causal Debiasing Medical Multimodal Representation Learning with Missing Modalities
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2509.05615