March 22, 2024Cancer Research

Multimodal Machine Learning for Cancer Type and Gleason Score Prediction in TCGA

View Full Paper

Ask AI

Bookmark

Authors

CWChristian WohlfartRoche Pharma AG (Germany)EKEldad KlaimanLithuanian University of Health Sciences JWJacub Witkowski

Discussion

Member takes

Implication

Machine learning study demonstrates improved diagnostic accuracy in cancer cohorts from TCGA, suggesting multimodal data fusion enhances tumor characterization.

Key Points

Multimodal machine learning combining histopathology images and RNA-seq enhances classification accuracy across diverse human cancer types, outperforming unimodal predictive models.
The integrated framework achieved an F1 score of 0.95 for pancancer classification and 0.73 for prostate Gleason score grading in test validation cohorts.
Assessment using matched whole slide images and RNA-seq merged linear SVM and Resnet18 probabilities, highlighting problem-specific utility across oncology applications.

Cite This Study

Wohlfart et al. (2024) studied this question.

synapsesocial.com/papers/68e72ceab6db6435876a7165 https://doi.org/10.1158/1538-7445.am2024-4970

View Full Paper

Ask AI

Bookmark

Implication

Machine learning study demonstrates improved diagnostic accuracy in cancer cohorts from TCGA, suggesting multimodal data fusion enhances tumor characterization.

Multimodal Machine Learning for Cancer Type and Gleason Score Prediction in TCGA

Discussion

Member takes

Implication

Key Points

Cite This Study

Multimodal Machine Learning for Cancer Type and Gleason Score Prediction in TCGA

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider