Computer-Aided Diagnosis of Liver Cancers Using Deep Learning with Fine-tuning

• Item type: SIG Technical Reports(1)
• 公開日: 2018-09-13
• タイトル: Computer-Aided Diagnosis of Liver Cancers Using Deep Learning with Fine-tuning
• タイトル / 言語: en
• タイトル: Computer-Aided Diagnosis of Liver Cancers Using Deep Learning with Fine-tuning
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: ディスカッションセッション4
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Ritsumeikan University
• 著者所属: Zhejiang University
• 著者所属: Zhejiang University
• 著者所属: Zhejiang University
• 著者所属: Zhejiang University
• 著者所属: Zhejiang University
• 著者所属: Ritsumeikan University
• 著者所属: Ritsumeikan University
• 著者所属: Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): Zhejiang University
• 著者所属(英): en
• 著者所属(英): Zhejiang University
• 著者所属(英): en
• 著者所属(英): Zhejiang University
• 著者所属(英): en
• 著者所属(英): Zhejiang University
• 著者所属(英): en
• 著者所属(英): Zhejiang University
• 著者所属(英): en
• 著者所属(英): Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): Ritsumeikan University
• 著者名: Weibin, Wang; Dong, Liang; Lanfen, Lin; Hongjie, Hu; Qiaowei, Zhang; Qingqing, Chen; Yutaro, Iwamoto; Xianhua, Han; Yen-Wei, Chen
• 著者名(英): Weibin, Wang; Dong, Liang; Lanfen, Lin; Hongjie, Hu; Qiaowei, Zhang; Qingqing, Chen; Yutaro, Iwamoto; Xianhua, Han; Yen-Wei, Chen
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Liver cancer is one of the leading causes of death world-wide. Computer-aided diagnosis plays an important role in liver lesion diagnosis (classification). Recently, several deep learning-based computer-aided diagnosis systems have been proposed for classification of liver lesions and their effectiveness have been demonstrated. The main challenge in deep learningbased medical image classification is the lack of annotated training samples. In this paper, we demonstrated that fine-tuning can significantly improve the liver lesion classification accuracy especially for the small training samples. We used the residual convolutional neural network (ResNet), which is the state-of-the-art network, as our baseline network for focal liver lesion classification on multi-phase CT images. The fine-tuning significantly improved the classification accuracy from 83.7% to 91.2%. The classification accuracy (91.2%) is higher than the accuracy of the state-of-the-art methods.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Liver cancer is one of the leading causes of death world-wide. Computer-aided diagnosis plays an important role in liver lesion diagnosis (classification). Recently, several deep learning-based computer-aided diagnosis systems have been proposed for classification of liver lesions and their effectiveness have been demonstrated. The main challenge in deep learningbased medical image classification is the lack of annotated training samples. In this paper, we demonstrated that fine-tuning can significantly improve the liver lesion classification accuracy especially for the small training samples. We used the residual convolutional neural network (ResNet), which is the state-of-the-art network, as our baseline network for focal liver lesion classification on multi-phase CT images. The fine-tuning significantly improved the classification accuracy from 83.7% to 91.2%. The classification accuracy (91.2%) is higher than the accuracy of the state-of-the-art methods.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AA11131797
• 書誌情報: 研究報告コンピュータビジョンとイメージメディア（CVIM）; 巻 2018-CVIM-213, 号 23, p. 1-2, 発行日 2018-09-13
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8701
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会