研究者詳細 - 森　健策

論文 - 森　健策

分割表示 1,775 件中 61 - 80 件目／全件表示 >>

Attention-Guided Pancreatic Duct Segmentation from Abdominal CT Volumes 招待有り査読有り

Shen C., Roth H.R., Yuichiro H., Oda M., Miyamoto T., Sato G., Mori K.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 12969 LNCS 巻頁： 46 - 55 2021年

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Pancreatic duct dilation indicates a high risk of pancreatic ductal adenocarcinoma (PDAC), the deadliest cancer with a poor prognosis. Segmentation of dilated pancreatic duct from CT taken from patients without PDAC shows the potential to assist the early detection of PDAC. Most current researches include pancreatic duct segmentation as one additional class for patients who have already detected PDAC. However, the dilated pancreatic duct for people who have not yet developed PDAC is typically much smaller, making the segmentation difficult. Deep learning-based segmentation on tiny components is challenging because of the large imbalance between the target object and irrelevant regions. In this work, we explore an attention-guided approach for dilated pancreatic duct segmentation as a screening tool for pre-PDAC patients, enhancing the pancreas regions’ concentration and ignoring the unnecessary features. We employ a multi-scale aggregation to combine the information at different scales to improve the segmentation performance further. Our proposed multi-scale pancreatic attention-guided approach achieved a Dice score of 54.16% on dilated pancreatic duct dataset, which shows a significant improvement over prior techniques.

DOI： 10.1007/978-3-030-90874-4_5

Scopus

Current status and future perspective on artificial intelligence for lower endoscopy 招待有り査読有り

Misawa M., Kudo S.E., Mori Y., Maeda Y., Ogawa Y., Ichimasa K., Kudo T., Wakamura K., Hayashi T., Miyachi H., Baba T., Ishida F., Itoh H., Oda M., Mori K.

Gastroenterological Endoscopy 63 巻 ( 7 ) 頁： 1402 - 1416 2021年

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Gastroenterological Endoscopy

The global incidence and mortality rate of colorectal cancer remains high. Colonoscopy is regarded as the gold standard examination for detecting and eradicating neoplastic lesions. However, there are some uncertainties in colonoscopy practice that are related to limitations in human performance. First, approximately one-fourth of colorectal neoplasms are missed on a single colonoscopy. Second, it is still difficult for nonexperts to perform adequately regarding optical biopsy. Third, recording of some quality indicators (e.g. cecal intubation, bowel preparation, and withdrawal speed）which are related to adenoma detection rate, is sometimes incomplete. With recent improvements in machine learning techniques and advances in computer performance, artificial intelligence-assisted computer-aided diagnosis is being increasingly utilized by endoscopists. In particular, the emergence of deep-learning, data-driven machine learning techniques have made the development of computer-aided systems easier than that of conventional machine learning techniques, the former currently being considered the standard artificial intelligence engine of computer-aided diagnosis by colonoscopy. To date, computer-aided detection systems seem to have improved the rate of detection of neoplasms. Additionally, computer-aided characterization systems may have the potential to improve diagnostic accuracy in real-time clinical practice. Furthermore, some artificial intelligence-assisted systems that aim to improve the quality of colonoscopy have been reported. The implementation of computer-aided system clinical practice may provide additional benefits such as helping in educational poorly performing endoscopists and supporting real-time clinical decision-making. In this review, we have focused on computer-aided diagnosis during colonoscopy reported by gastroenterologists and discussed its status, limitations, and future prospects.

DOI： 10.11280/gee.63.1402

Scopus

CiNii Research

1．機械学習によるCOVID-19症例CT画像の診断支援招待有り査読有り

森健策

映像情報メディア学会誌 75 巻 ( 3 ) 頁： 325 - 329 2021年

Clinical and Genetic Characteristics of 18 Patients from 13 Japanese Families with CRX-associated retinal disorder: Identification of Genotype-phenotype Association 招待有り査読有り

Fujinami-Yokokawa Y., Fujinami K., Kuniyoshi K., Hayashi T., Ueno S., Mizota A., Shinoda K., Arno G., Pontikos N., Yang L., Liu X., Sakuramoto H., Katagiri S., Mizobuchi K., Kominami T., Terasaki H., Nakamura N., Kameya S., Yoshitake K., Miyake Y., Kurihara T., Tsubota K., Miyata H., Iwata T., Tsunoda K., Nishimura T., Hayashizaki Y., Kondo M., Shimozawa N., Horiguchi M., Yamamoto S., Kuze M., Naoi N., Machida S., Shimada Y., Nakamura M., Fujikado T., Hotta Y., Takahashi M., Mochizuki K., Murakami A., Kondo H., Ishida S., Nakazawa M., Hatase T., Matsunaga T., Maeda A., Noda K., Tanikawa A., Yamamoto S., Yamamoto H., Araie M., Aihara M., Nakazawa T., Sekiryu T., Kashiwagi K., Kosaki K., Piero C., Fukuchi T., Hayashi A., Hosono K., Mori K., Tanaka K., Furuya K., Suzuki K., Kohata R., Yanagi Y., Minegishi Y., Iejima D., Suga A., Rossmiller B.P., Pan Y., Oshima T., Nakayama M., Teruyama Y., Yamamoto M., Minematsu N., Sanbe H., Mori D., Kijima Y., Mawatari G., Kurata K., Yamada N., Itoh M., Kawaji H., Murakawa Y.

Scientific Reports 10 巻 ( 1 ) 2020年12月

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Scientific Reports

Inherited retinal disorder (IRD) is a leading cause of blindness, and CRX is one of a number of genes reported to harbour autosomal dominant (AD) and recessive (AR) causative variants. Eighteen patients from 13 families with CRX-associated retinal disorder (CRX-RD) were identified from 730 Japanese families with IRD. Ophthalmological examinations and phenotype subgroup classification were performed. The median age of onset/latest examination was 45.0/62.5 years (range, 15–77/25–94). The median visual acuity in the right/left eye was 0.52/0.40 (range, −0.08–2.00/−0.18–1.70) logarithm of the minimum angle of resolution (LogMAR) units. There was one family with macular dystrophy, nine with cone-rod dystrophy (CORD), and three with retinitis pigmentosa. In silico analysis of CRX variants was conducted for genotype subgroup classification based on inheritance and the presence of truncating variants. Eight pathogenic CRX variants were identified, including three novel heterozygous variants (p.R43H, p.P145Lfs*42, and p.P197Afs*22). A trend of a genotype-phenotype association was revealed between the phenotype and genotype subgroups. A considerably high proportion of CRX-RD in ADCORD was determined in the Japanese cohort (39.1%), often showing the mild phenotype (CORD) with late-onset disease (sixth decade). Frequently found heterozygous missense variants located within the homeodomain underlie this mild phenotype. This large cohort study delineates the disease spectrum of CRX-RD in the Japanese population.

DOI： 10.1038/s41598-020-65737-z

Scopus

Clinical characteristics in patients with ossification of the posterior longitudinal ligament: A prospective multi-institutional cross-sectional study 招待有り査読有り

Hirai T., Yoshii T., Ushio S., Mori K., Maki S., Katsumi K., Nagoshi N., Takeuchi K., Furuya T., Watanabe K., Nishida N., Watanabe K., Kaito T., Kato S., Nagashima K., Koda M., Ito K., Imagama S., Matsuoka Y., Wada K., Kimura A., Ohba T., Katoh H., Matsuyama Y., Ozawa H., Haro H., Takeshita K., Watanabe M., Matsumoto M., Nakamura M., Yamazaki M., Okawa A., Kawaguchi Y.

Scientific Reports 10 巻 ( 1 ) 2020年12月

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Scientific Reports

Ossification of the posterior longitudinal ligament (OPLL) can occur throughout the entire spine and can sometimes lead to spinal disorder. Although patients with OPLL sometimes develop physical limitations because of pain, the characteristics of pain and effects on activities of daily living (ADL) have not been precisely evaluated in OPLL patients. Therefore, we conducted a multi-center prospective study to assess whether the symptoms of cervical OPLL are different from those of cervical spondylosis (CS). A total of 263 patients with a diagnosis of cervical OPLL and 50 patients with a diagnosis of CS were enrolled and provided self-reported outcomes, including responses to the Japanese Orthopaedic Association (JOA) Cervical Myelopathy Evaluation Questionnaire (JOACMEQ), JOA Back Pain Evaluation Questionnaire (JOABPEQ), visual analog scale (VAS), and SF-36 scores. The severity of myelopathy was significantly correlated with each domain of the JOACMEQ and JOABPEQ. There was a negative correlation between the VAS score for each domain and the JOA score. There were significantly positive correlations between the JOA score and the Mental Health, Bodily Pain, Physical Functioning, Role Emotional, and Role Physical domains of the SF-36. One-to-one matching resulted in 50 pairs of patients with OPLL and CS. Although there was no significant between-group difference in scores in any of the domains of the JOACMEQ or JOABPEQ, the VAS scores for pain or numbness in the buttocks or limbs were significantly higher in the CS group; however, there was no marked difference in low back pain, chest tightness, or numbness below the chest between the two study groups. The scores for the Role Physical and Body Pain domains of the SF-36 were significantly higher in the OPLL group than in the CS group, and the mean scores for the other domains was similar between the two groups. The results of this study revealed that patients with OPLL were likely to have neck and low back pain and restriction in ADL. No specific type of pain was found in patients with OPLL when compared with those who had CS.

DOI： 10.1038/s41598-020-62278-3

Scopus

Synthetic laparoscopic video generation for machine learning-based surgical instrument segmentation from real laparoscopic video and virtual surgical instruments 査読有り国際共著

Takuya Ozawa,Yuichiro Hayashi,Hirohisa Oda, Masahiro Oda, Takayuki Kitasaka, Nobuyoshi Takeshita, Masaaki Ito, Kensaku Mori

Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization 2020年11月

Clinical impact of Endoscopic Surgical Skill Qualification System (ESSQS) by Japan Society for Endoscopic Surgery (JSES) for laparoscopic distal gastrectomy and low anterior resection based on the National Clinical Database (NCD) registry. 招待有り査読有り

Akagi T, Endo H, Inomata M, Yamamoto H, Mori T, Kojima K, Kuroyanagi H, Sakai Y, Nakajima K, Shiroshita H, Etoh T, Saida Y, Yamamoto S, Hasegawa H, Ueno H, Kakeji Y, Miyata H, Kitagawa Y, Watanabe M

Annals of gastroenterological surgery 4 巻 ( 6 ) 頁： 721 - 734 2020年11月

A visual SLAM-based bronchoscope tracking scheme for bronchoscopic navigation 招待有り査読有り

Wang, C; Oda, M; Hayashi, Y; Villard, B; Kitasaka, T; Takabatake, H; Mori, M; Honma, H; Natori, H; Mori, K

INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY 15 巻 ( 10 ) 頁： 1619 - 1630 2020年10月

担当区分：責任著者記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：International Journal of Computer Assisted Radiology and Surgery

Purpose: Due to the complex anatomical structure of bronchi and the resembling inner surfaces of airway lumina, bronchoscopic examinations require additional 3D navigational information to assist the physicians. A bronchoscopic navigation system provides the position of the endoscope in CT images with augmented anatomical information. To overcome the shortcomings of previous navigation systems, we propose using a technique known as visual simultaneous localization and mapping (SLAM) to improve bronchoscope tracking in navigation systems. Methods: We propose an improved version of the visual SLAM algorithm and use it to estimate nt-specific bronchoscopic video as input. We improve the tracking procedure by adding more narrow criteria in feature matching to avoid mismatches. For validation, we collected several trials of bronchoscopic videos with a bronchoscope camera by exploring synthetic rubber bronchus phantoms. We simulated breath by adding periodic force to deform the phantom. We compared the camera positions from visual SLAM with the manually created ground truth of the camera pose. The number of successfully tracked frames was also compared between the original SLAM and the proposed method. Results: We successfully tracked 29,559 frames at a speed of 80 ms per frame. This corresponds to 78.1% of all acquired frames. The average root mean square error for our technique was 3.02 mm, while that for the original was 3.61 mm. Conclusion: We present a novel methodology using visual SLAM for bronchoscope tracking. Our experimental results showed that it is feasible to use visual SLAM for the estimation of the bronchoscope camera pose during bronchoscopic navigation. Our proposed method tracked more frames and showed higher accuracy than the original technique did. Future work will include combining the tracking results with virtual bronchoscopy and validation with in vivo cases.

DOI： 10.1007/s11548-020-02241-9

Web of Science

Scopus

PubMed

Station number assignment to abdominal lymph node for assisting gastric cancer surgery 査読有り

Yuichiro Hayashi, Kazunari Misawa, Kensaku Mori

Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization 2020年10月

A deformable model for navigated laparoscopic gastrectomy based on finite elemental method 招待有り査読有り

Chen Tao, Wei Guodong, Xu Lili, Shi Weili, Xu Yikai, Zhu Yongyi, Hayashi Yuichiro, Oda Hirohisa, Oda Masahiro, Hu Yanfeng, Yu Jiang, Jiang Zhengang, Li Guoxin, Mori Kensaku

MINIMALLY INVASIVE THERAPY & ALLIED TECHNOLOGIES 29 巻 ( 4 ) 頁： 210 - 216 2020年7月

担当区分：責任著者記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：Minimally Invasive Therapy and Allied Technologies

Background: Accurate registration for surgical navigation of laparoscopic surgery is highly challenging due to vessel deformation. Here, we describe the design of a deformable model with improved matching accuracy by applying the finite element method (FEM). Material and methods: ANSYS software was used to simulate an FEM model of the vessel after pull-up based on laparoscopic gastrectomy requirements. The central line of the FEM model and the central line of the ground truth were drawn and compared. Based on the material and parameters determined from the animal experiment, a perigastric vessel FEM model of a gastric cancer patient was created, and its accuracy in a laparoscopic gastrectomy surgical scene was evaluated. Results: In the animal experiment, the FEM model created with Ogden foam material exhibited better results. The average distance between the two central lines was 6.5mm, and the average distance between their closest points was 3.8 mm. In the laparoscopic gastrectomy surgical scene, the FEM model and the true artery deformation demonstrated good coincidence. Conclusion: In this study, a deformable vessel model based on FEM was constructed using preoperative CT images to improve matching accuracy and to supply a reference for further research on deformation matching to facilitate laparoscopic gastrectomy navigation.

DOI： 10.1080/13645706.2019.1625926

Web of Science

Scopus

PubMed

Automated laparoscopic colorectal surgery workflow recognition using artificial intelligence: Experimental research 招待有り査読有り

Kitaguchi Daichi, Takeshita Nobuyoshi, Matsuzaki Hiroki, Oda Tatsuya, Watanabe Masahiko, Mori Kensaku, Kobayashi Etsuko, Ito Masaaki

INTERNATIONAL JOURNAL OF SURGERY 79 巻頁： 88 - 94 2020年7月

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：International Journal of Surgery

Background: Identifying laparoscopic surgical videos using artificial intelligence (AI) facilitates the automation of several currently time-consuming manual processes, including video analysis, indexing, and video-based skill assessment. This study aimed to construct a large annotated dataset comprising laparoscopic colorectal surgery (LCRS) videos from multiple institutions and evaluate the accuracy of automatic recognition for surgical phase, action, and tool by combining this dataset with AI. Materials and methods: A total of 300 intraoperative videos were collected from 19 high-volume centers. A series of surgical workflows were classified into 9 phases and 3 actions, and the area of 5 tools were assigned by painting. More than 82 million frames were annotated for a phase and action classification task, and 4000 frames were annotated for a tool segmentation task. Of these frames, 80% were used for the training dataset and 20% for the test dataset. A convolutional neural network (CNN) was used to analyze the videos. Intersection over union (IoU) was used as the evaluation metric for tool recognition. Results: The overall accuracies for the automatic surgical phase and action classification task were 81.0% and 83.2%, respectively. The mean IoU for the automatic tool segmentation task for 5 tools was 51.2%. Conclusions: A large annotated dataset of LCRS videos was constructed, and the phase, action, and tool were recognized with high accuracy using AI. Our dataset has potential uses in medical applications such as automatic video indexing and surgical skill assessments. Open research will assist in improving CNN models by making our dataset available in the field of computer vision.

DOI： 10.1016/j.ijsu.2020.05.015

Web of Science

Scopus

Artificial intelligence for magnifying endoscopy, endocytoscopy, and confocal laser endomicroscopy of the colorectum 招待有り査読有り

Mori, Y; Kudo, SE; Misawa, M; Itoh, H; Oda, M; Mori, K

TECHNIQUES AND INNOVATIONS IN GASTROINTESTINAL ENDOSCOPY 22 巻 ( 2 ) 頁： 56 - 60 2020年4月

担当区分：責任著者記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：Techniques and Innovations in Gastrointestinal Endoscopy

Because magnifying endoscopy is considered to be more accurate at predicting the histology of colorectal polyps than nonmagnifying endoscopy, it has been attracting a lot of attention, especially in Japan. However, use of magnifying endoscopy is not yet widespread because of its limited availability and the difficulty in interpreting the acquired images. Application of artificial intelligence (AI) is now changing this situation because it helps less-skilled endoscopists to accurately interpret magnified images. Research in this field initially focused on magnifying endoscopy with narrow-band imaging as the target of AI. Most previously published retrospective studies have reported over 90% sensitivity in differentiation of neoplastic lesions; however, automatically indicating the region of interest (ROI) of the polyps that AI should analyze has been found to be challenging. To address this practical problem, some researchers have started to adopt contact endomicroscopy as a target for AI. Contact endomicroscopy includes endocytoscopy (520-fold magnification, Olympus, Tokyo, Japan) and confocal laser endomicroscopy (1000-fold magnification, Mauna Kea, Paris, France). These forms of contact endomicroscopy provide ultramagnified images that make it unnecessary to manually select the ROI because the entire image acquired by contact endomicroscopy is the ROI of the targeted polyps. This strength of contact endomicroscopy has contributed to early implementation of this technology into clinical practice, which may change the utility of magnifying endoscopy in clinical settings and help increase its use globally in the near future.

DOI： 10.1016/j.tgie.2019.150632

Web of Science

Scopus

Clinical application of a surgical navigation system based on virtual thoracoscopy for lung cancer patients: real time visualization of area of lung cancer before induction therapy and optimal resection line for obtaining a safe surgical margin during surgery 招待有り査読有り

Nakamura Shota, Hayashi Yuichiro, Kawaguchi Koji, Fukui Takayuki, Hakiri Shuhei, Ozeki Naoki, Mori Shunsuke, Goto Masaki, Mori Kensaku, Yokoi Kohei

JOURNAL OF THORACIC DISEASE 12 巻 ( 3 ) 頁： 672 - 679 2020年3月

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Journal of Thoracic Disease

Background: We have developed a surgical navigation system that presents virtual thoracoscopic images using computed tomography (CT) image data, as if you are observing intra-thoracic cavity in synchronization with the real thoracoscopic view. Using this system, we made it possible to simultaneously visualize the ‘area of lung cancer before induction therapy’ and the ‘optimal resection line for obtaining a safe surgical margin’ as a virtual thoracoscopic view. We applied this navigation system in the clinical setting in operations for lung cancer patients with chest wall invasion after induction chemoradiotherapy. Methods: The proposed surgical navigation system consisted of a three-dimensional (3D) positional tracker and a virtual thoracoscopy system. The 3D positional tracker was used to recognize the positional information of the real thoracoscope. The virtual thoracoscopy system generated virtual thoracoscopic views based on CT image data. Combined with these two technologies, patient-to-image registration was performed in two patients, and the results generated a virtual thoracoscopic view that was synchronized with the real thoracoscopic view. Results: The operations were started with video-assisted thoracic surgery (VATS), and the navigation system was activated at the same time. The virtual thoracoscopic view was synchronized with the real thoracoscopic view, which also simultaneously indicated the ‘area of lung cancer before induction therapy’ and the ‘optimal resection lines for obtaining a safe surgical margin’. We marked the optimal lines using an electric scalpel, and then performed lobectomy and chest wall resection with a sufficient surgical margin using these landmarks. Pathological examinations confirmed that the surgical margin was negative. No complications related to the navigation system were encountered during or after the procedures. Conclusions: Using this proposed navigation system, we could obtain a ‘CT-derived virtual intrathoracic 3D view of the patient’ that was aligned with the thoracoscopic view during surgery. The accurate identification of areas of cancer invasion before induction therapy using this system might be a useful for determining optimal surgical resection lines.

DOI： 10.21037/jtd.2019.12.108

Web of Science

Scopus

Tensor-cut: A tensor-based graph-cut blood vessel segmentation method and its application to renal artery segmentation 査読有り国際共著

Chenglong Wang, Masahiro Oda, Yuichiro Hayashi, Yasushi Yoshino, Tokunori Yamamoto, Alejandro F. Frangi, Kensaku Mori

Medical Image Analysis 60 巻頁： 101623 2020年2月

3Dプリンティングの最新動向招待有り

森健策

インナービジョン 35 巻頁： 36 - 37 2020年

[Pharmacological action and clinical effect of tedizolid phosphate (SIVEXTRO<sup>®</sup> Tablets 200 mg, for iv infusion 200 mg), a novel oxazolidinone-class antibacterial drug]. 招待有り査読有り

Mori M, Takase A

Nihon yakurigaku zasshi. Folia pharmacologica Japonica 155 巻 ( 5 ) 頁： 332 - 339 2020年

Automated Pancreas Segmentation Using Multi-institutional Collaborative Deep Learning 招待有り査読有り

Wang P., Shen C., Roth H.R., Yang D., Xu D., Oda M., Misawa K., Chen P.T., Liu K.L., Liao W.C., Wang W., Mori K.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 12444 LNCS 巻頁： 192 - 200 2020年

The performance of deep learning based methods strongly relies on the number of datasets used for training. Many efforts have been made to increase the data in the medical image analysis field. However, unlike photography images, it is hard to generate centralized databases to collect medical images because of numerous technical, legal, and privacy issues. In this work, we study the use of federated learning between two institutions in a real-world setting to collaboratively train a model without sharing the raw data across national boundaries. We quantitatively compare the segmentation models obtained with federated learning and local training alone. Our experimental results show that federated learning models have higher generalizability than standalone training.

DOI： 10.1007/978-3-030-60548-3_19

Scopus

Automated eye disease classification method from anterior eye image using anatomical structure focused image classification technique 招待有り査読有り

Oda Masahiro, Yamaguchi Takefumi, Fukuoka Hideki, Ueno Yuta, Mori Kensaku

MEDICAL IMAGING 2020: COMPUTER-AIDED DIAGNOSIS 11314 巻 2020年

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Progress in Biomedical Optics and Imaging - Proceedings of SPIE

This paper presents an automated classification method of infective and non-infective diseases from anterior eye images. Treatments for cases of infective and non-infective diseases are different. Distinguishing them from anterior eye images is important to decide a treatment plan. Ophthalmologists distinguish them empirically. Quantitative classification of them based on computer assistance is necessary. We propose an automated classification method of anterior eye images into cases of infective or non-infective disease. Anterior eye images have large variations of the eye position and brightness of illumination. This makes the classification difficult. If we focus on the cornea, positions of opacified areas in the corneas are different between cases of the infective and non-infective diseases. Therefore, we solve the anterior eye image classification task by using an object detection approach targeting the cornea. This approach can be said as "anatomical structure focused image classification". We use the YOLOv3 object detection method to detect corneas of infective disease and corneas of non-infective disease. The detection result is used to define a classification result of an image. In our experiments using anterior eye images, 88.3% of images were correctly classified by the proposed method.

DOI： 10.1117/12.2549951

Web of Science

Scopus

2. 多元計算解剖モデルと人工知能に基づく診断治療支援招待有り

森健策

映像情報メディア学会誌 74 巻 ( 6 ) 頁： 909 - 915 2020年

Abdominal artery segmentation method from CT volumes using fully convolutional neural network 招待有り査読有り

Oda, M; Roth, HR; Kitasaka, T; Misawa, K; Fujiwara, M; Mori, K

INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY 14 巻 ( 12 ) 頁： 2069 - 2081 2019年12月

担当区分：責任著者記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：International Journal of Computer Assisted Radiology and Surgery

Purpose : The purpose of this paper is to present a fully automated abdominal artery segmentation method from a CT volume. Three-dimensional (3D) blood vessel structure information is important for diagnosis and treatment. Information about blood vessels (including arteries) can be used in patient-specific surgical planning and intra-operative navigation. Since blood vessels have large inter-patient variations in branching patterns and positions, a patient-specific blood vessel segmentation method is necessary. Even though deep learning-based segmentation methods provide good segmentation accuracy among large organs, small organs such as blood vessels are not well segmented. We propose a deep learning-based abdominal artery segmentation method from a CT volume. Because the artery is one of small organs that is difficult to segment, we introduced an original training sample generation method and a three-plane segmentation approach to improve segmentation accuracy. Method : Our proposed method segments abdominal arteries from an abdominal CT volume with a fully convolutional network (FCN). To segment small arteries, we employ a 2D patch-based segmentation method and an area imbalance reduced training patch generation (AIRTPG) method. AIRTPG adjusts patch number imbalances between patches with artery regions and patches without them. These methods improved the segmentation accuracies of small artery regions. Furthermore, we introduced a three-plane segmentation approach to obtain clear 3D segmentation results from 2D patch-based processes. In the three-plane approach, we performed three segmentation processes using patches generated on axial, coronal, and sagittal planes and combined the results to generate a 3D segmentation result. Results : The evaluation results of the proposed method using 20 cases of abdominal CT volumes show that the averaged F-measure, precision, and recall rates were 87.1%, 85.8%, and 88.4%, respectively. This result outperformed our previous automated FCN-based segmentation method. Our method offers competitive performance compared to the previous blood vessel segmentation methods from 3D volumes. Conclusions : We developed an abdominal artery segmentation method using FCN. The 2D patch-based and AIRTPG methods effectively segmented the artery regions. In addition, the three-plane approach generated good 3D segmentation results.

DOI： 10.1007/s11548-019-02062-5

<森 健策>

論文 - 森 健策

<森　健策>

論文 - 森　健策