3D-Texture-Segmentation of Prostate Cancer from Multimodal MRI Data

Kraisorn Chaisaowong, Markus Kitza

international conference on electrical engineering/electronics, computer, telecommunications and information technology(2021)

引用 0|浏览0
暂无评分
摘要
Prostate cancer is the fourth most commonly diagnosed cancer worldwide. The TRUS-guided prostate biopsy e.g. is painful and lack accuracy. A noninvasive, reproducible and precise localization of prostate cancer without causing pain is still required. Multispectral MRI data of various modalities are currently promising technique for prostate cancer localization. However, it is obvious that these 2D images are difficult to read. With the help of a computerized segmentation the time required to read the 3D images could be reduced and the accuracy of eventual biopsies could be increased by focusing on a specific region. For this purpose, three major 3D textural features are exploited, i.e. GLCM, LBP, and three dimensional wavelet transform, together with T1-weighted, T2-weighted and contrast enhanced data will be used to generate a multispectral basis for the features. In this paper, all mentioned enhanced MRI features are investigated, at the end the best promising 3D features are tested to classify the prostate cancer area.
更多
查看译文
关键词
Prostate Tumor,Multispectral MRI,Multimodal MRI,3D-Texture,Feature Selection,Classification
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要