期刊
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
卷 33, 期 2, 页码 729-745出版社
WILEY
DOI: 10.1002/ima.22816
关键词
computed tomography; contrast enhancement; K-means clustering; liver tumor segmentation; power-law transformation; SegNet; semantic segmentation
This article presents a deep learning-based method for liver segmentation, which accurately segments the liver and liver tumors. By using the SegNet model and the K-means clustering method combined with image enhancement techniques, good segmentation results were obtained, and the method demonstrated stability across all phases.
Liver and liver tumor segmentations are essential in computer-aided systems for diagnosing liver tumors. These systems must operate on multiphase computed tomography (CT) images instead of a single phase for accurate diagnosis for clinical applications. We have proposed a framework that can perform segmentation from quadriphasic CT data. The liver was segmented using a fine-tuned SegNet model and the liver tumor was segmented using the K-means clustering method coupled with a power-law transformation-based image enhancement technique. The best values for liver segmentation achieved were: Dice Coefficient = 96.46 +/- 0.48%, Jaccard Index = 93.16 +/- 0.89%, volumetric overlap error = 6.84 +/- 0.89% and average symmetric surface distance = 0.59 +/- 0.3 mm and the results for liver tumor delineation were Dice Coefficient = 85.07 +/- 4.5%, Jaccard Index = 74.29 +/- 6.8%, volumetric overlap error = 25.71 +/- 6.8% and average symmetric surface distance = 1.14 +/- 1.3 mm. The proposed liver segmentation method based on deep learning is fully automatic, robust, and effective for all phases. The image enhancement technique has shown promising results and aided in better liver tumor segmentation. The liver tumors were segmented satisfactorily; however, improvements concerning false positive reduction can further increase the accuracy.
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