期刊
METHODS
卷 208, 期 -, 页码 48-58出版社
ACADEMIC PRESS INC ELSEVIER SCIENCE
DOI: 10.1016/j.ymeth.2022.10.005
关键词
Medical image segmentation; Whole heart segmentation; Deep forest; Cardiac CT image segmentation
资金
- Natural Science Foundation of Fujian Province of China [2020J01006]
- National Natural Science Foundation of China [61502402]
- Open Project Program of State Key Laboratory of Virtual Reality Technology and Systems, Beihang University [VRLAB2022AC04]
- University Distin- guished Young Research Talent Training Program of Fujian Province
- Hong Kong Innovation Technology Commission
This paper proposes an improved Deep Forest framework (MRDFF) for automatic whole heart segmentation. The framework consists of two stages, where the first stage extracts the heart region through binary classification and the second stage subdivides the results to obtain accurate cardiac substructures. Additionally, methods such as feature fusion, multi-resolution fusion, and multi-scale fusion are proposed to further improve segmentation accuracy.
Automatic whole heart segmentation plays an important role in the treatment and research of cardiovascular diseases. In this paper, we propose an improved Deep Forest framework, named Multi-Resolution Deep Forest Framework (MRDFF), which accomplishes whole heart segmentation in two stages. We extract the heart region by binary classification in the first stage, thus avoiding the class imbalance problem caused by too much background. The results of the first stage are then subdivided in the second stage to obtain accurate cardiac substructures. In addition, we also propose hybrid feature fusion, multi-resolution fusion and multi-scale fusion to further improve the segmentation accuracy. Experiments on the public dataset MM-WHS show that our model can achieve comparable accuracy in about half the training time of neural network models.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据