Universal and Extensible Language-Vision Models for Organ Segmentation and Tumor Detection from Abdominal Computed Tomography

TL;DR

This paper introduces a universal, extensible AI framework for organ segmentation and tumor detection in CT scans, leveraging language embeddings and class-specific heads to improve flexibility, accuracy, and efficiency across diverse datasets.

Contribution

The paper proposes a novel language-driven parameter generator and class-specific heads, enabling a single model to handle multiple datasets, adapt to new classes, and outperform existing methods.

Findings

Achieves top performance on six CT segmentation tasks in MSD leaderboard.

Demonstrates 6x faster inference compared to dataset-specific models.

Shows strong generalization and transferability across different datasets.

Abstract

The advancement of artificial intelligence (AI) for organ segmentation and tumor detection is propelled by the growing availability of computed tomography (CT) datasets with detailed, per-voxel annotations. However, these AI models often struggle with flexibility for partially annotated datasets and extensibility for new classes due to limitations in the one-hot encoding, architectural design, and learning scheme. To overcome these limitations, we propose a universal, extensible framework enabling a single model, termed Universal Model, to deal with multiple public datasets and adapt to new classes (e.g., organs/tumors). Firstly, we introduce a novel language-driven parameter generator that leverages language embeddings from large language models, enriching semantic encoding compared with one-hot encoding. Secondly, the conventional output layers are replaced with lightweight, class-specific heads, allowing Universal Model to simultaneously segment 25 organs and six types of tumors and ease the addition of new classes. We train our Universal Model on 3,410 CT volumes assembled from 14 publicly available datasets and then test it on 6,173 CT volumes from four external datasets. Universal Model achieves first place on six CT tasks in the Medical Segmentation Decathlon (MSD) public leaderboard and leading performance on the Beyond The Cranial Vault (BTCV) dataset. In summary, Universal Model exhibits remarkable computational efficiency (6x faster than other dataset-specific models), demonstrates strong generalization across different hospitals, transfers well to numerous downstream tasks, and more importantly, facilitates the extensibility to new classes while alleviating the catastrophic forgetting of previously learned classes. Codes, models, and datasets are available at https://github.com/ljwztc/CLIP-Driven-Universal-Model