arXiv:2506.05127v2 Announce Type: replace-cross 
Abstract: The digitization of histology slides has revolutionized pathology, providing massive datasets for cancer diagnosis and research. Self-supervised and vision-language models have been shown to effectively mine large pathology datasets to learn discriminative representations. On the other hand, there are unique problems in pathology, such as annotated data scarcity, privacy regulations in data sharing, and inherently generative tasks like virtual staining. Generative models, capable of synthesizing realistic and diverse images, present a compelling solution to address these problems through image synthesis. We introduce PixCell, the first generative foundation model for histopathology images. PixCell is a diffusion model trained on PanCan-30M, a large, diverse dataset derived from 69,184 H&E-stained whole slide images of various cancer types. We employ a progressive training strategy and a self-supervision-based conditioning that allows us to scale up training without any human-annotated data. By conditioning on real slides, the synthetic images capture the properties of the real data and can be used as data augmentation for small-scale datasets to boost classification performance. We prove the foundational versatility of PixCell by applying it to two generative downstream tasks: privacy-preserving synthetic data generation and virtual IHC staining. PixCell's high-fidelity conditional generation enables institutions to use their private data to synthesize highly realistic, site-specific surrogate images that can be shared in place of raw patient data. Furthermore, using datasets of roughly paired H&E-IHC tiles, we learn to translate PixCell's conditioning from H&E to multiple IHC stains, allowing the generation of IHC images from H&E inputs. Our trained models are publicly released to accelerate research in computational pathology.

تم تقديم PixCell كأول نموذج أساسي توليدي مصمم خصيصًا لصور علم الأمراض الرقمية، باستخدام نموذج انتشار تم تدريبه على مجموعة بيانات PanCan-30M الضخمة، التي تتكون من 69,184 صورة شريحة كاملة ملونة بـ H&E من أنواع مختلفة من السرطان.

PixCell se ha presentado como el primer modelo de fundación generativo diseñado específicamente para imágenes de histopatología digital, utilizando un modelo de difusión entrenado en el extenso conjunto de datos PanCan-30M, que comprende 69,184 imágenes de diapositivas completas teñidas con H&E de varios tipos de cáncer.

PixCell a été introduit comme le premier modèle de fondation génératif spécifiquement conçu pour les images d'histopathologie numériques, utilisant un modèle de diffusion entraîné sur l'énorme ensemble de données PanCan-30M, qui comprend 69 184 images de diapositives entières colorées à l'H&E provenant de divers types de cancer.

PixCell has been introduced as the first generative foundation model specifically designed for digital histopathology images, utilizing a diffusion model trained on the extensive PanCan-30M dataset, which comprises 69,184 H&E-stained whole slide images from various cancer types.

PixCell: A generative foundation model for digital histopathology images

One More Thing in AI – Your Shortcut to AI Mastery

PixCell: A generative foundation model for digital histopathology images

Was this article worth reading? Share it

One More Thing in AI

4o Image Gen

Republiclabs.ai

ChartPixel

Pixel Dojo

Komiko

Ready to build your own newsroom?