Probing Intrinsic Medical Task Relationships: A Contrastive Learning Perspective

2026-04-07

AI Summary

Existing research has not systematically uncovered the intrinsic relationships among medical vision tasks at the representation level. This work proposes TaCo, a task contrastive learning framework that, for the first time, maps 30 heterogeneous medical vision tasks—including segmentation, detection, image generation, and transformation—into a unified embedding space from a data-driven perspective. Leveraging 39 datasets spanning multiple imaging modalities, TaCo constructs a comprehensive task relational atlas. By introducing task embeddings and a contrastive learning mechanism, the framework effectively captures structural relationships across tasks and modalities, revealing both clustering tendencies and continuous variation patterns among tasks. These insights provide a theoretical foundation and empirical evidence for the design of multitask medical vision models.

Abstract

While much of the medical computer vision community has focused on advancing performance for specific tasks, the underlying relationships between tasks, i.e., how they relate, overlap, or differ on a representational level, remain largely unexplored. Our work explores these intrinsic relationships between medical vision tasks, specifically, we investigate 30 tasks, such as semantic tasks (e.g., segmentation and detection), image generative tasks (e.g., denoising, inpainting, or colorization), and image transformation tasks (e.g., geometric transformations). Our goal is to probe whether a data-driven representation space can capture an underlying structure of tasks across a variety of 39 datasets from wildly different medical imaging modalities, including computed tomography, magnetic resonance, electron microscopy, X-ray ultrasound and more. By revealing how tasks relate to one another, we aim to provide insights into their fundamental properties and interconnectedness. To this end, we introduce Task-Contrastive Learning (TaCo), a contrastive learning framework designed to embed tasks into a shared representation space. Through TaCo, we map these heterogeneous tasks from different modalities into a joint space and analyze their properties: identifying which tasks are distinctly represented, which blend together, and how iterative alterations to tasks are reflected in the embedding space. Our work provides a foundation for understanding the intrinsic structure of medical vision tasks, offering a deeper understanding of task similarities and their interconnected properties in embedding spaces.

Problem

• medical vision tasks
• task relationships
• representation space
• contrastive learning
• medical imaging modalities

Innovation

• Task-Contrastive Learning
• medical vision tasks
• representation space
• task relationships
• contrastive learning