A GAN conditioned on contrastive-aligned histopathology images and clinical text generates gene expression profiles, with attention maps showing which tissue regions drove the synthesis.
Bioinformatics39(Supplement_1), i111–i120 (06 2023)
1 Pith paper cite this work, alongside 34 external citations. Polarity classification is still indexing.
1
Pith paper citing it
34
external citations · OpenAlex
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
M$^3$-Gen: Interpretable Multimodal Generation of Gene Expression Profiles Using Clinical and Imaging Data
A GAN conditioned on contrastive-aligned histopathology images and clinical text generates gene expression profiles, with attention maps showing which tissue regions drove the synthesis.