VALSE 论文速览 第205期：监督式3D模型在医疗图像分析任务中的应用 ...

论文题目：

How well do supervised models transfer to 3d image segmentation?

作者列表：

B站观看网址：

https://www.bilibili.com/video/BV1BDxNeeEuB/?spm_id_from=333.999.0.0&vd_source=2044ac986b5caa4c8b5f00b525441df3

复制链接到浏览器打开或点击阅读原文即可跳转至观看页面。

论文摘要：

The pre-training and fine-tuning paradigm has become prominent in transfer learning. For example, if the model is pre-trained on ImageNet and then fine-tuned to PASCAL, it can significantly outperform that trained on PASCAL from scratch. While ImageNet pre-training has shown enormous success, it is formed in 2D and the learned features are for classification tasks; when transferring to more diverse tasks, like 3D image segmentation, its performance is inevitably compromised due to the deviation from the original ImageNet context. A significant challenge lies in the lack of large, annotated 3D datasets rivaling the scale of ImageNet for model pre-training. To overcome this challenge, we make two contributions. Firstly, we construct AbdomenAtlas 1.1 that comprises 9,262 three-dimensional computed tomography (CT) volumes with high-quality, per-voxel annotations of 25 anatomical structures and pseudo annotations of seven tumor types. Secondly, we develop a suite of models that are pre-trained on our AbdomenAtlas 1.1 for transfer learning. Our preliminary analyses indicate that the model trained only with 21 CT volumes, 672 masks, and 40 GPU hours has a transfer learning ability similar to the model trained with 5,050 (unlabeled) CT volumes and 1,152 GPU hours. More importantly, the transfer learning ability of supervised models can further scale up with larger annotated datasets, achieving significantly better performance than preexisting pre-trained models, irrespective of their pre-training methodologies or data sources. We hope this study can facilitate collective efforts in constructing larger 3D vision datasets and more releases of supervised pre-trained models.

参考文献：

[1] Li, Wenxuan, Alan Yuille, and Zongwei Zhou. "How well do supervised models transfer to 3d image segmentation?." In The Twelfth International Conference on Learning Representations. 2024.

论文链接：

[https://www.cs.jhu.edu/~alanlab/Pubs23/li2023transitioning.pdf]

代码链接：

[https://github.com/MrGiovanni/SuPreM]

视频讲者简介：

Wenxuan Li, a student of Computer Science from Johns Hopkins University, is mentored by Professor Alan Yuille and Dr. Zongwei Zhou. Her research focuses on utilizing large-scale datasets and annotations to enhance medical image analysis. She has published several papers in top conferences such as ICLR and RSNA. She has also served as a leader for organizing several challenges, including ISBI and MICCAI.

个人主页：

https://github.com/MrGiovanni/SuPreM/tree/main

特别鸣谢本次论文速览主要组织者：

月度轮值AC：武宇 (武汉大学)