.png){kind=icon}
返回
Advancing 3D medical image analysis with variable dimension transform based supervised 3D pre-training
.png){kind=icon}
OA
AI
DOI：10.1016/j.neucom.2023.01.012.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
The difficulties in both data acquisition and annotation substantially restrict the sample sizes of training datasets for 3D medical imaging applications. Therefore, it is non-trivial to build well-performing 3D con-volutional neural networks from scratch. Previous efforts on 3D pre-training have frequently relied on self-supervised approaches, which use either predictive or contrastive learning on unlabeled data to build invariant 3D representations. However, because of the unavailability of large-scale supervision informa-tion, obtaining semantically invariant and discriminative representations from these learning frame-works remains problematic. In this paper, we revisit an innovative yet simple fully-supervised 3D network pre-training framework to take advantage of semantic supervision from large-scale 2D natural image datasets. With a redesigned 3D network architecture, reformulated natural images are used to address the problem of data scarcity and develop powerful 3D representations. Comprehensive experi-ments on five benchmark datasets demonstrate that the proposed pre-trained models can effectively accelerate convergence while also improving accuracy for a variety of 3D medical imaging tasks such as classification, segmentation, and detection. In addition, as compared to training from scratch, it can save up to 60% of annotation efforts. On the NIH DeepLesion dataset, it also achieves state-of-the-art detection performance, outperforming earlier self-supervised and fully-supervised pre-training approaches, as well as methods that do training from scratch. To facilitate further development of 3D medical models, our code and pre-trained model weights are publicly available at https://github.com/u rmagicsmine/CSPR. (c) 2023 Elsevier B.V. All rights reserved.
Keyword:
3D medical image
Transfer learning
Variable dimension transform
Supervised pre-training
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
6.5
论文数:
2.5W
被引数:
6.5W
机构
引用论文
Prognostic Value of Deep Learning PET/CT-Based Radiomics: Potential Role for Future Individual Induction Chemotherapy in Advanced Nasopharyngeal Carcinoma基于深度学习PET/CT的影像组学的预后价值: 晚期鼻咽癌未来个体化诱导化疗的潜在作用.png)
CLINICAL CANCER RESEARCHOAAI
CLINICAL CANCER RESEARCHIF10.2
OAAI分享收藏Automatic Segmentation of Acute Ischemic Stroke From DWI Using 3-D Fully Convolutional DenseNets使用3-D完全卷积的DWI自动分割急性缺血性卒中
PREAI
分享收藏U-Net: deep learning for cell counting, detection, and morphometryU-net: 用于细胞计数、检测和形态测量的深度学习NATURE METHODS
NATURE METHODSIF32.1
PREAI
分享收藏The Lung Image Database Consortium, (LIDC) and Image Database Resource Initiative (IDRI): A Completed Reference Database of Lung Nodules on CT Scans肺图像数据库联盟 (LIDC) 和图像数据库资源倡议 (IDRI): ct扫描上完整的肺结节参考数据库MEDICAL PHYSICS
OAAI
MEDICAL PHYSICSIF3.2
OAAI分享收藏Self-supervised learning for medical image analysis using image context restorationMEDICAL IMAGE ANALYSISOAAI
MEDICAL IMAGE ANALYSISIF11.8
OAAI分享收藏分享收藏Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs用于检测视网膜眼底照片中糖尿病视网膜病变的深度学习算法的开发和验证
PREAI
分享收藏
查看更多内容