Robust multi-modal prostate cancer classification via feature autoencoder and dual attention
Prostate cancer is the second leading cause of cancer death in men. At present, the methods for classifying early cancer grades on MRI images are mainly focused on single image modality and with low robustness. Therefore, this paper focuses on exploring the method of classifying cancer grades on mul...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-01-01
|
| Series: | Informatics in Medicine Unlocked |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2352914822000727 |
| _version_ | 1818009268541980672 |
|---|---|
| author | Bochong Li Ryo Oka, M.D Ping Xuan Yuichiro Yoshimura, PhD Toshiya Nakaguchi |
| author_facet | Bochong Li Ryo Oka, M.D Ping Xuan Yuichiro Yoshimura, PhD Toshiya Nakaguchi |
| author_sort | Bochong Li |
| collection | DOAJ |
| description | Prostate cancer is the second leading cause of cancer death in men. At present, the methods for classifying early cancer grades on MRI images are mainly focused on single image modality and with low robustness. Therefore, this paper focuses on exploring the method of classifying cancer grades on multi-modality MRI images and maintaining robustness. In this paper, we propose a novel and effective multi-modal convolutional neural network for discriminating prostate cancer clinical severity grade, i.e., Robust Multi-modal Feature Autoencoder Attention net (RMANet); this model greatly improves the accuracy and robustness of the model. T2-weighted and Diffusion-weighted imaging are used in this article. The model consists of two branches, one of them is to learn the overall features of two MRI modalities by building a ten-layer CNN network with two input shared weights, and the other branch uses auto-encoder structure with classical U-net as the backbone to learn specific features of each modality and to improve the robustness of the classification model. In the branch of learning overall features of each modality, the novel dual attention mechanism is added to this branch, through which the attention mechanism can better direct the learning focus of the model to the cancerous regions. Experiments were conducted on the ProstateX dataset and augmented with hospital data. By comparing with other baseline methods, multi-modal input methods, and State-of-the-Art (SOTA) methods, the AUC values obtained by the proposed model (reaching 0.84) in this paper after the test set are higher than other classical models and most recent methods, and the sensitivity values (reaching 0.84) are higher than the recent method. |
| first_indexed | 2024-04-14T05:40:47Z |
| format | Article |
| id | doaj.art-f96237d4532d40069529870b11460f2d |
| institution | Directory Open Access Journal |
| issn | 2352-9148 |
| language | English |
| last_indexed | 2024-04-14T05:40:47Z |
| publishDate | 2022-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Informatics in Medicine Unlocked |
| spelling | doaj.art-f96237d4532d40069529870b11460f2d2022-12-22T02:09:29ZengElsevierInformatics in Medicine Unlocked2352-91482022-01-0130100923Robust multi-modal prostate cancer classification via feature autoencoder and dual attentionBochong Li0Ryo Oka, M.D1Ping Xuan2Yuichiro Yoshimura, PhD3Toshiya Nakaguchi4Graduate School of Science and Technology, Chiba University, 1-33, Yayoicho, Inage Ward, Chiba-shi, 263-002, Japan; Corresponding author.Toho University Sakura Medical Center, 564-1 Shimoshizu, Sakura, Chiba, Sakura, 285-0841, JapanSchool of Computer Science and Technology, Heilongjiang University, 74 Xuefu Road, Nangang District, Harbin, Harbin, 150080, ChinaToyama University School of Medicine, 3190 Gofuku, Toyama, Gofuku, 930-8555, JapanCenter for Frontier Medical Engineering, Chiba University, 1-33, Yayoicho, Inage Ward, Chiba-shi, Chiba, 263-8522, JapanProstate cancer is the second leading cause of cancer death in men. At present, the methods for classifying early cancer grades on MRI images are mainly focused on single image modality and with low robustness. Therefore, this paper focuses on exploring the method of classifying cancer grades on multi-modality MRI images and maintaining robustness. In this paper, we propose a novel and effective multi-modal convolutional neural network for discriminating prostate cancer clinical severity grade, i.e., Robust Multi-modal Feature Autoencoder Attention net (RMANet); this model greatly improves the accuracy and robustness of the model. T2-weighted and Diffusion-weighted imaging are used in this article. The model consists of two branches, one of them is to learn the overall features of two MRI modalities by building a ten-layer CNN network with two input shared weights, and the other branch uses auto-encoder structure with classical U-net as the backbone to learn specific features of each modality and to improve the robustness of the classification model. In the branch of learning overall features of each modality, the novel dual attention mechanism is added to this branch, through which the attention mechanism can better direct the learning focus of the model to the cancerous regions. Experiments were conducted on the ProstateX dataset and augmented with hospital data. By comparing with other baseline methods, multi-modal input methods, and State-of-the-Art (SOTA) methods, the AUC values obtained by the proposed model (reaching 0.84) in this paper after the test set are higher than other classical models and most recent methods, and the sensitivity values (reaching 0.84) are higher than the recent method.http://www.sciencedirect.com/science/article/pii/S2352914822000727Prostate cancerComputer-aided detectionMagnetic resonance imagingMachine learning |
| spellingShingle | Bochong Li Ryo Oka, M.D Ping Xuan Yuichiro Yoshimura, PhD Toshiya Nakaguchi Robust multi-modal prostate cancer classification via feature autoencoder and dual attention Informatics in Medicine Unlocked Prostate cancer Computer-aided detection Magnetic resonance imaging Machine learning |
| title | Robust multi-modal prostate cancer classification via feature autoencoder and dual attention |
| title_full | Robust multi-modal prostate cancer classification via feature autoencoder and dual attention |
| title_fullStr | Robust multi-modal prostate cancer classification via feature autoencoder and dual attention |
| title_full_unstemmed | Robust multi-modal prostate cancer classification via feature autoencoder and dual attention |
| title_short | Robust multi-modal prostate cancer classification via feature autoencoder and dual attention |
| title_sort | robust multi modal prostate cancer classification via feature autoencoder and dual attention |
| topic | Prostate cancer Computer-aided detection Magnetic resonance imaging Machine learning |
| url | http://www.sciencedirect.com/science/article/pii/S2352914822000727 |
| work_keys_str_mv | AT bochongli robustmultimodalprostatecancerclassificationviafeatureautoencoderanddualattention AT ryookamd robustmultimodalprostatecancerclassificationviafeatureautoencoderanddualattention AT pingxuan robustmultimodalprostatecancerclassificationviafeatureautoencoderanddualattention AT yuichiroyoshimuraphd robustmultimodalprostatecancerclassificationviafeatureautoencoderanddualattention AT toshiyanakaguchi robustmultimodalprostatecancerclassificationviafeatureautoencoderanddualattention |