Beyond Pixel Simulation: Pathology Image Generation via Diagnostic Semantic Tokens and Prototype Control

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Han, Minghao, Liu, Yichen, Liu, Yizhou, Chen, Zizhi, Tang, Jingqun, Wu, Xuecheng, Yang, Dingkang, Zhang, Lihua
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912927383027712
author Han, Minghao
Liu, Yichen
Liu, Yizhou
Chen, Zizhi
Tang, Jingqun
Wu, Xuecheng
Yang, Dingkang
Zhang, Lihua
author_facet Han, Minghao
Liu, Yichen
Liu, Yizhou
Chen, Zizhi
Tang, Jingqun
Wu, Xuecheng
Yang, Dingkang
Zhang, Lihua
contents In computational pathology, understanding and generation have evolved along disparate paths: advanced understanding models already exhibit diagnostic-level competence, whereas generative models largely simulate pixels. Progress remains hindered by three coupled factors: the scarcity of large, high-quality image-text corpora; the lack of precise, fine-grained semantic control, which forces reliance on non-semantic cues; and terminological heterogeneity, where diverse phrasings for the same diagnostic concept impede reliable text conditioning. We introduce UniPath, a semantics-driven pathology image generation framework that leverages mature diagnostic understanding to enable controllable generation. UniPath implements Multi-Stream Control: a Raw-Text stream; a High-Level Semantics stream that uses learnable queries to a frozen pathology MLLM to distill paraphrase-robust Diagnostic Semantic Tokens and to expand prompts into diagnosis-aware attribute bundles; and a Prototype stream that affords component-level morphological control via a prototype bank. On the data front, we curate a 2.65M image-text corpus and a finely annotated, high-quality 68K subset to alleviate data scarcity. For a comprehensive assessment, we establish a four-tier evaluation hierarchy tailored to pathology. Extensive experiments demonstrate UniPath's SOTA performance, including a Patho-FID of 80.9 (51% better than the second-best) and fine-grained semantic control achieving 98.7% of the real-image. The dataset and code can be obtained from https://github.com/Hanminghao/UniPath.
format Preprint
id arxiv_https___arxiv_org_abs_2512_21058
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Beyond Pixel Simulation: Pathology Image Generation via Diagnostic Semantic Tokens and Prototype Control
Han, Minghao
Liu, Yichen
Liu, Yizhou
Chen, Zizhi
Tang, Jingqun
Wu, Xuecheng
Yang, Dingkang
Zhang, Lihua
Computer Vision and Pattern Recognition
In computational pathology, understanding and generation have evolved along disparate paths: advanced understanding models already exhibit diagnostic-level competence, whereas generative models largely simulate pixels. Progress remains hindered by three coupled factors: the scarcity of large, high-quality image-text corpora; the lack of precise, fine-grained semantic control, which forces reliance on non-semantic cues; and terminological heterogeneity, where diverse phrasings for the same diagnostic concept impede reliable text conditioning. We introduce UniPath, a semantics-driven pathology image generation framework that leverages mature diagnostic understanding to enable controllable generation. UniPath implements Multi-Stream Control: a Raw-Text stream; a High-Level Semantics stream that uses learnable queries to a frozen pathology MLLM to distill paraphrase-robust Diagnostic Semantic Tokens and to expand prompts into diagnosis-aware attribute bundles; and a Prototype stream that affords component-level morphological control via a prototype bank. On the data front, we curate a 2.65M image-text corpus and a finely annotated, high-quality 68K subset to alleviate data scarcity. For a comprehensive assessment, we establish a four-tier evaluation hierarchy tailored to pathology. Extensive experiments demonstrate UniPath's SOTA performance, including a Patho-FID of 80.9 (51% better than the second-best) and fine-grained semantic control achieving 98.7% of the real-image. The dataset and code can be obtained from https://github.com/Hanminghao/UniPath.
title Beyond Pixel Simulation: Pathology Image Generation via Diagnostic Semantic Tokens and Prototype Control
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.21058