Generating crossmodal gene expression from cancer histopathology improves multimodal AI predictions

Samiran Dey, Christopher R.S. Banerji, Partha Basuchowdhuri, Sanjoy K. Saha, Deepak Parashar, Tapabrata Chakraborti · 2025

Read Generating crossmodal gene expression from cancer histopathology improves multimodal AI predictions by Samiran Dey, Christopher R.S. Banerji, Partha Basuchowdhuri, Sanjoy K. Saha, Deepak Parashar, Tapabrata Chakraborti (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-00568.txt.

Description

Samiran Dey and five co-authors show that gene expression made from tumor images can stand in for real molecular tests, in a two thousand twenty-five medical paper. Joining pathology slides with transcriptomics improves cancer grading and survival prediction, but tests of gene activity are rarely ordered in public healthcare.

Their diffusion-based model, PathGen, builds gene expression profiles straight from tissue-slide images. Across four cohorts spanning glioma, renal, uterine, and breast cancers, adding made-up data greatly improved grading and risk estimates. Calls with synthetic features could not be told apart from those using real transcriptomics. The system also reports well-scaled uncertainty and attention maps that doctors and nurses can inspect.

Catalog record

CREATORSamiran Dey, Christopher R.S. Banerji, Partha Basuchowdhuri, Sanjoy K. Saha, Deepak Parashar, Tapabrata Chakraborti
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE88 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-00568

Browse the full library · About the Tower of Babel