ProFound: A moderate-sized vision foundation model for multi-task prostate imaging

2026-03-04

AI Summary

This work addresses the challenge of automating clinical tasks in multiparametric MRI (mpMRI) for prostate cancer, which is typically hindered by the need for large amounts of task-specific annotated data. To this end, we develop a medium-scale vision foundation model tailored for prostate mpMRI, pretrained on over 22,000 3D MRI volumes from 5,000 patients using diverse self-supervised learning strategies to capture generalizable imaging representations. Our model is the first in prostate imaging to unify a single architecture across eleven heterogeneous clinical tasks, substantially reducing reliance on labeled data. Evaluation on an independent cohort of more than 3,000 patients demonstrates that, after fine-tuning, the model achieves or surpasses the performance of current state-of-the-art task-specific models in key applications including cancer detection, Gleason grading, and lesion localization.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Other Foundations and Applications

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Many diagnostic and therapeutic clinical tasks for prostate cancer increasingly rely on multi-parametric MRI. Automating these tasks is challenging because they necessitate expert interpretations, which are difficult to scale to capitalise on modern deep learning. Although modern automated systems achieve expert-level performance in isolated tasks, their general clinical utility remains limited by the requirement of large task-specific labelled datasets. In this paper, we present ProFound, a domain-specialised vision foundation model for volumetric prostate mpMRI. ProFound is pre-trained using several variants of self-supervised approaches on a diverse, multi-institutional collection of 5,000 patients, with a total of over 22,000 unique 3D MRI volumes (over 1,800,000 2D image slices). We conducted a systematic evaluation of ProFound across a broad spectrum of $11$ downstream clinical tasks on over 3,000 independent patients, including prostate cancer detection, Gleason grading, lesion localisation, gland volume estimation, zonal and surrounding structure segmentation. Experimental results demonstrate that finetuned ProFound consistently outperforms or remains competitive with state-of-the-art specialised models and existing medical vision foundation models trained/finetuned on the same data.

Problem

• prostate cancer
• multi-parametric MRI
• vision foundation model
• clinical task automation
• medical image analysis

Innovation

• vision foundation model
• self-supervised learning
• prostate mpMRI
• multi-task learning
• medical image analysis