Baseline Method of the Foundation Model Challenge for Ultrasound Image Analysis

2026-02-01

AI Summary

This study addresses the challenge of developing a generalizable and clinically deployable foundation model for ultrasound imaging, which is hindered by substantial anatomical and acquisition protocol variability. To this end, the authors propose a unified multi-head multi-task learning framework (MH-MTL) that jointly handles 27 diverse tasks—including segmentation, classification, detection, and regression—within a single shared network. Built upon an ImageNet-pretrained EfficientNet-B4 backbone and a feature pyramid network (FPN) for multi-scale feature extraction, the framework incorporates task-adaptive learning rate scaling and cosine annealing. A task routing mechanism enables differentiated fusion of high-level semantics and spatial details, significantly enhancing cross-task generalization. This work establishes the first unified multi-task foundation model benchmark for ultrasound image analysis, demonstrating its feasibility and robustness on the FM_UIA~2026 dataset and providing a strong, extensible baseline for future research.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Ultrasound (US) imaging exhibits substantial heterogeneity across anatomical structures and acquisition protocols, posing significant challenges to the development of generalizable analysis models. Most existing methods are task-specific, limiting their suitability as clinically deployable foundation models. To address this limitation, the Foundation Model Challenge for Ultrasound Image Analysis (FM\_UIA~2026) introduces a large-scale multi-task benchmark comprising 27 subtasks across segmentation, classification, detection, and regression. In this paper, we present the official baseline for FM\_UIA~2026 based on a unified Multi-Head Multi-Task Learning (MH-MTL) framework that supports all tasks within a single shared network. The model employs an ImageNet-pretrained EfficientNet--B4 backbone for robust feature extraction, combined with a Feature Pyramid Network (FPN) to capture multi-scale contextual information. A task-specific routing strategy enables global tasks to leverage high-level semantic features, while dense prediction tasks exploit spatially detailed FPN representations. Training incorporates a composite loss with task-adaptive learning rate scaling and a cosine annealing schedule. Validation results demonstrate the feasibility and robustness of this unified design, establishing a strong and extensible baseline for ultrasound foundation model research. The code and dataset are publicly available at \href{https://github.com/lijiake2408/Foundation-Model-Challenge-for-Ultrasound-Image-Analysis}{GitHub}.

Problem

• ultrasound image analysis
• foundation model
• task-specific models
• heterogeneity
• generalizable models

Innovation

• Multi-Task Learning
• Foundation Model
• Ultrasound Image Analysis
• Feature Pyramid Network
• Task-Specific Routing