HiCAF-Net: A Hierarchical Cross-Attention Fusion framework for cross-cancer subtype classification using histopathological and genomic data

Junyi Wu

doi:10.1016/j.compmedimag.2026.102788

HiCAF-Net: A Hierarchical Cross-Attention Fusion framework for cross-cancer subtype classification using histopathological and genomic data

2026-06-10

0

PRE

AI

J

Junyi Wu

C

Chenyu Zhao

J

Jiaqi Yuan

Q

Qingyin Zhou

Y

Yilin Wei

J

Jianmin Li

E

Eugene Edzeafene-Mensah

O

Ouchen Wang

C

Chenhui Yang

M

Meihao Wang

Z

Zhifang Pan *

DOI：10.1016/j.compmedimag.2026.102788

Original

Original request for help

Save

Abstract

En 中文

• Unified framework analyzes eight heterogeneous cancer types. • Hierarchical fusion captures pan-cancer visual hallmarks. • Collaborative optimization prevents negative task transfer. • Domain adaptation bridges shifts across fragmented data.

HiCAF-Net: A Hierarchical Cross-Attention Fusion framework for cross-cancer subtype classification using histopathological and genomic data

Abstract

Journal

Organization

Cited Papers

Citing Papers