HiLa: Hierarchical Vision-Language Collaboration for Cancer Survival Prediction

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Cui, Jiaqi, Wen, Lu, Fei, Yuchen, Liu, Bo, Zhou, Luping, Shen, Dinggang, Wang, Yan
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913929623502848
author Cui, Jiaqi
Wen, Lu
Fei, Yuchen
Liu, Bo
Zhou, Luping
Shen, Dinggang
Wang, Yan
author_facet Cui, Jiaqi
Wen, Lu
Fei, Yuchen
Liu, Bo
Zhou, Luping
Shen, Dinggang
Wang, Yan
contents Survival prediction using whole-slide images (WSIs) is crucial in cancer re-search. Despite notable success, existing approaches are limited by their reliance on sparse slide-level labels, which hinders the learning of discriminative repre-sentations from gigapixel WSIs. Recently, vision language (VL) models, which incorporate additional language supervision, have emerged as a promising solu-tion. However, VL-based survival prediction remains largely unexplored due to two key challenges. First, current methods often rely on only one simple lan-guage prompt and basic cosine similarity, which fails to learn fine-grained associ-ations between multi-faceted linguistic information and visual features within WSI, resulting in inadequate vision-language alignment. Second, these methods primarily exploit patch-level information, overlooking the intrinsic hierarchy of WSIs and their interactions, causing ineffective modeling of hierarchical interac-tions. To tackle these problems, we propose a novel Hierarchical vision-Language collaboration (HiLa) framework for improved survival prediction. Specifically, HiLa employs pretrained feature extractors to generate hierarchical visual features from WSIs at both patch and region levels. At each level, a series of language prompts describing various survival-related attributes are constructed and aligned with visual features via Optimal Prompt Learning (OPL). This ap-proach enables the comprehensive learning of discriminative visual features cor-responding to different survival-related attributes from prompts, thereby improv-ing vision-language alignment. Furthermore, we introduce two modules, i.e., Cross-Level Propagation (CLP) and Mutual Contrastive Learning (MCL) to maximize hierarchical cooperation by promoting interactions and consistency be-tween patch and region levels. Experiments on three TCGA datasets demonstrate our SOTA performance.
format Preprint
id arxiv_https___arxiv_org_abs_2507_04613
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HiLa: Hierarchical Vision-Language Collaboration for Cancer Survival Prediction
Cui, Jiaqi
Wen, Lu
Fei, Yuchen
Liu, Bo
Zhou, Luping
Shen, Dinggang
Wang, Yan
Computer Vision and Pattern Recognition
Artificial Intelligence
Survival prediction using whole-slide images (WSIs) is crucial in cancer re-search. Despite notable success, existing approaches are limited by their reliance on sparse slide-level labels, which hinders the learning of discriminative repre-sentations from gigapixel WSIs. Recently, vision language (VL) models, which incorporate additional language supervision, have emerged as a promising solu-tion. However, VL-based survival prediction remains largely unexplored due to two key challenges. First, current methods often rely on only one simple lan-guage prompt and basic cosine similarity, which fails to learn fine-grained associ-ations between multi-faceted linguistic information and visual features within WSI, resulting in inadequate vision-language alignment. Second, these methods primarily exploit patch-level information, overlooking the intrinsic hierarchy of WSIs and their interactions, causing ineffective modeling of hierarchical interac-tions. To tackle these problems, we propose a novel Hierarchical vision-Language collaboration (HiLa) framework for improved survival prediction. Specifically, HiLa employs pretrained feature extractors to generate hierarchical visual features from WSIs at both patch and region levels. At each level, a series of language prompts describing various survival-related attributes are constructed and aligned with visual features via Optimal Prompt Learning (OPL). This ap-proach enables the comprehensive learning of discriminative visual features cor-responding to different survival-related attributes from prompts, thereby improv-ing vision-language alignment. Furthermore, we introduce two modules, i.e., Cross-Level Propagation (CLP) and Mutual Contrastive Learning (MCL) to maximize hierarchical cooperation by promoting interactions and consistency be-tween patch and region levels. Experiments on three TCGA datasets demonstrate our SOTA performance.
title HiLa: Hierarchical Vision-Language Collaboration for Cancer Survival Prediction
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2507.04613