用分阶段扩散模型统一实现多模态多器官病灶分割，效果超越现有方法。
UniSegDiff: Boosting Unified Lesion Segmentation via a Staged Diffusion Model
- 分阶段训练与推理，动态调整不同阶段的预测目标。
- 在六种器官、多种模态上均显著优于当前最优方法。
- 适合医学图像分割研究者，尤其关注跨模态统一建模场景。
扩散概率模型（DPM）在各类生成任务中表现出色。其固有的随机性有助于缓解医学图像和标签边缘模糊的问题，使DPM成为病灶分割的有前景方法。然而我们发现，当前扩散模型的训练与推理策略导致不同时刻注意力分布不均，造成训练时间长且解不优。为此，我们提出UniSegDiff，一种新颖的扩散模型框架，旨在统一处理多模态、多器官的病灶分割。该框架采用分阶段训练与推理策略，动态调整各阶段预测目标，强制模型在所有时刻保持高注意力，并通过预训练分割特征提取网络实现统一分割。我们在六种不同器官、多种成像模态上评估性能。实验结果表明，UniSegDiff显著优于此前的最先进方法。代码已开源：https://github.com/HUYILONG-Z/UniSegDiff。
原文摘要 · Abstract (English)
The Diffusion Probabilistic Model (DPM) has demonstrated remarkable performance across a variety of generative tasks. The inherent randomness in diffusion models helps address issues such as blurring at the edges of medical images and labels, positioning Diffusion Probabilistic Models (DPMs) as a promising approach for lesion segmentation. However, we find that the current training and inference strategies of diffusion models result in an uneven distribution of attention across different timesteps, leading to longer training times and suboptimal solutions. To this end, we propose UniSegDiff, a novel diffusion model framework designed to address lesion segmentation in a unified manner across multiple modalities and organs. This framework introduces a staged training and inference approach, dynamically adjusting the prediction targets at different stages, forcing the model to maintain high attention across all timesteps, and achieves unified lesion segmentation through pre-training the feature extraction network for segmentation. We evaluate performance on six different organs across various imaging modalities. Comprehensive experimental results demonstrate that UniSegDiff significantly outperforms previous state-of-the-art (SOTA) approaches. The code is available at https://github.com/HUYILONG-Z/UniSegDiff.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。