Learning Comprehensive Multimodal Representation for Cancer Survival Prediction
Open Access
- 23 December 2022
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 332-336
Abstract
Making accurate cancer survival prediction plays a key role in helping clinicians to plan appropriate treatments, which can enhance the life quality of cancer patients and reduce the rate of cancer death. Recent works demonstrate that combining information from different modalities, e.g., genomic data and histopathology images, is crucial for improving the performance of cancer survival prediction. Although these works have obtained promising results, the heterogeneity of data distributions poses a great challenge to efficiently learning comprehensive multimodal representations. To solve this problem, we propose a novel method, which learns both modality-invariant and -specific representations to provide a holistic view of multimodal data. Experimental results suggest that our method achieves remarkable performance and outperforms other approaches.Keywords
Funding Information
- 61871361, 61971393, 61471331, 61571414
This publication has 13 references indexed in Scilit:
- Pathomic Fusion: An Integrated Framework for Fusing Histopathology and Genomic Features for Cancer Diagnosis and PrognosisIEEE Transactions on Medical Imaging, 2022
- Deep Semisupervised Multitask Learning Model and Its Interpretability for Survival AnalysisIEEE Journal of Biomedical and Health Informatics, 2021
- Deep learning with multimodal representation for pancancer prognosis predictionBioinformatics, 2019
- Integrative Analysis of Pathological Images and Multi-Dimensional Genomic Data for Early-Stage Cancer PrognosisIEEE Transactions on Medical Imaging, 2019
- Integrating genomic data and pathological images to effectively predict breast cancer clinical outcomeComputer Methods and Programs in Biomedicine, 2018
- Predicting cancer outcomes from histology and genomics using convolutional networksProceedings of the National Academy of Sciences of the United States of America, 2018
- Integrative Analysis of Histopathological Images and Genomic Data Predicts Clear Cell Renal Cell Carcinoma PrognosisCancer Research, 2017
- Predicting non-small cell lung cancer prognosis by fully automated microscopic pathology image featuresNature Communications, 2016
- Open Access to Large Scale Datasets Is Needed to Translate Knowledge of Cancer Heterogeneity into Better Patient OutcomesPLOS Medicine, 2015
- Quantitative Image Analysis of Cellular Heterogeneity in Breast Tumors Complements Genomic ProfilingScience Translational Medicine, 2012