Unified Modeling Enhanced Multimodal Learning for Precision Neuro-Oncology

Huahui Yi, Xiaofei Wang, Kang Li, Chao Li (Lead / Corresponding author)

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Multimodal learning, integrating histology images and genomics, promises to enhance precision oncology with comprehensive views at microscopic and molecular levels. However, existing methods may not sufficiently model the shared or complementary information for more effective integration. In this study, we introduce a Unified Modeling Enhanced Multimodal Learning (UMEML) framework that employs a hierarchical attention structure to effectively leverage shared and complementary features of both modalities of histology and genomics. Specifically, to mitigate unimodal bias from modality imbalance, we utilize a query-based cross-attention mechanism for prototype clustering in the pathology encoder. Our prototype assignment and modularity strategy are designed to align shared features and minimizes modality gaps. An additional registration mechanism with learnable tokens is introduced to enhance cross-modal feature integration and robustness in multimodal unified modeling. Our experiments demonstrate that our method surpasses previous state-of-the-art approaches in glioma diagnosis and prognosis tasks, underscoring its superiority in precision neuro-Oncology.

Original languageEnglish
Title of host publicationComputational Mathematics Modeling in Cancer Analysis
Subtitle of host publicationThird International Workshop, CMMCA 2024, Marrakesh, Morocco, October 6, 2024, Proceedings
EditorsJia Wu, Wenjian Qin, Chao Li, Boklye Kim
PublisherSpringer Science and Business Media Deutschland GmbH
Pages1-10
Number of pages10
Edition1
ISBN (Electronic)9783031733604
ISBN (Print)9783031733598
DOIs
Publication statusPublished - 5 Oct 2024
Event3rd Workshop on Computational Mathematics Modeling in Cancer Analysis - Marrakesh, Morocco
Duration: 6 Oct 20246 Oct 2024
https://cmmcaworkshop.github.io/2024/ (Link to Conference)

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume15181 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference3rd Workshop on Computational Mathematics Modeling in Cancer Analysis
Abbreviated titleCMMCA 2024
Country/TerritoryMorocco
CityMarrakesh
Period6/10/246/10/24
OtherCMMCA 2024 was held in conjunction with the 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024
Internet address

Keywords

  • Glioma
  • Multimodal classification
  • Multimodal learning
  • Survival prediction

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Unified Modeling Enhanced Multimodal Learning for Precision Neuro-Oncology'. Together they form a unique fingerprint.

Cite this