研究者詳細 - 森　健策

論文 - 森　健策

分割表示 1,841 件中 21 - 40 件目／全件表示 >>

Speedometer for withdrawal time monitoring during colonoscopy: a clinical implementation trial. 招待有り査読有り Open Access

Barua I, Misawa M, Glissen Brown JR, Walradt T, Kudo SE, Sheth SG, Nee J, Iturrino J, Mukherjee R, Cheney CP, Sawhney MS, Pleskow DK, Mori K, Løberg M, Kalager M, Wieszczy P, Bretthauer M, Berzin TM, Mori Y

Scandinavian journal of gastroenterology 58 巻 ( 6 ) 頁： 664 - 670 2023年6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：Scandinavian Journal of Gastroenterology

Objectives: Meticulous inspection of the mucosa during colonoscopy, represents a lengthier withdrawal time, but has been shown to increase adenoma detection rate (ADR). We investigated if artificial intelligence-aided speed monitoring can improve suboptimal withdrawal time. Methods: We evaluated the implementation of a computer-aided speed monitoring device during colonoscopy at a large academic endoscopy center. After informed consent, patients ≥18 years undergoing colonoscopy between 5 March and 29 April 2021 were examined without the use of the speedometer, and with the speedometer between 29 April and 30 June 2021. All colonoscopies were recorded, and withdrawal time was assessed based on the recordings in a blinded fashion. We compared mean withdrawal time, percentage of withdrawal time ≥6 min, and ADR with and without the speedometer. Results: One hundred sixty-six patients in each group were eligible for analyses. Mean withdrawal time was 9 min and 6.6 s (95% CI: 8 min and 34.8 s to 9 min and 39 s) without the use of the speedometer, and 9 min and 9 s (95% CI: 8 min and 45 s to 9 min and 33.6 s) with the speedometer; difference 2.3 s (95% CI: −42.3–37.7, p = 0.91). The ADRs were 45.2% (95% CI: 37.6–52.8) without the speedometer as compared to 45.8% (95% CI: 38.2–53.4) with the speedometer (p = 0.91). The proportion of colonoscopies with withdrawal time ≥6 min without the speedometer was 85.5% (95% CI: 80.2–90.9) versus 86.7% (95% CI: 81.6–91.9) with the speedometer (p = 0.75). Conclusions: Use of speed monitoring during withdrawal did not increase withdrawal time or ADR in colonoscopy. ClinicalTrials.gov Identifier: NCT04710251.

DOI： 10.1080/00365521.2022.2154616

Correction to: Gaussian affinity and GIoU-based loss for perforation detection and localization from colonoscopy videos. 招待有り査読有り Open Access

Jiang K, Itoh H, Oda M, Okumura T, Mori Y, Misawa M, Hayashi T, Kudo SE, Mori K

International journal of computer assisted radiology and surgery 18 巻 ( 5 ) 頁： 807 2023年5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

The original version of this article unfortunately contained a mistake. The incorrect notations were given in the author’s affiliations.

DOI： 10.1007/s11548-023-02899-x

Gaussian affinity and GIoU-based loss for perforation detection and localization from colonoscopy videos. 招待有り査読有り

Jiang K, Itoh H, Oda M, Okumura T, Mori Y, Misawa M, Hayashi T, Kudo SE, Mori K

International journal of computer assisted radiology and surgery 18 巻 ( 5 ) 頁： 795 - 805 2023年5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: Endoscopic submucosal dissection (ESD) is a minimally invasive treatment for early gastric cancer. However, perforations may happen and cause peritonitis during ESD. Thus, there is a potential demand for a computer-aided diagnosis system to support physicians in ESD. This paper presents a method to detect and localize perforations from colonoscopy videos to avoid perforation ignoring or enlarging by ESD physicians. Method: We proposed a training method for YOLOv3 by using GIoU and Gaussian affinity losses for perforation detection and localization in colonoscopic images. In this method, the object functional contains the generalized intersection over Union loss and Gaussian affinity loss. We propose a training method for the architecture of YOLOv3 with the presented loss functional to detect and localize perforations precisely. Results: To qualitatively and quantitatively evaluate the presented method, we created a dataset from 49 ESD videos. The results of the presented method on our dataset revealed a state-of-the-art performance of perforation detection and localization, which achieved 0.881 accuracy, 0.869 AUC, and 0.879 mean average precision. Furthermore, the presented method is able to detect a newly appeared perforation in 0.1 s. Conclusions: The experimental results demonstrated that YOLOv3 trained by the presented loss functional were very effective in perforation detection and localization. The presented method can quickly and precisely remind physicians of perforation happening in ESD. We believe a future CAD system can be constructed for clinical applications with the proposed method.

DOI： 10.1007/s11548-022-02821-x

Class-wise confidence-aware active learning for laparoscopic images segmentation 招待有り査読有り

Qiu, J; Hayashi, Y; Oda, M; Kitasaka, T; Mori, K

INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY 18 巻 ( 3 ) 頁： 473 - 482 2023年3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: Segmentation tasks are important for computer-assisted surgery systems as they provide the shapes of organs and the locations of instruments. What prevents the most powerful segmentation approaches from becoming practical applications is the requirement for annotated data. Active learning provides strategies to dynamically select the most informative samples to reduce the annotation workload. However, most previous active learning literature has failed to select the frames that containing low-appearing frequency classes, even though the existence of these classes is common in laparoscopic videos, resulting in poor performance in segmentation tasks. Furthermore, few previous works have explored the unselected data to improve active learning. Therefore, in this work, we focus on these classes to improve the segmentation performance. Methods: We propose a class-wise confidence bank that stores and updates the confidence scores for each class and a new acquisition function based on a confidence bank. We apply confidence scores to explore an unlabeled dataset by combining it with a class-wise data mixture method to exploit unlabeled datasets without any annotation. Results: We validated our proposal on two open-source datasets, CholecSeg8k and RobSeg2017, and observed that its performance surpassed previous active learning studies with about 10 % improvement on CholecSeg8k, especially for classes with a low-appearing frequency. For robSeg2017, we conducted experiments with a small and large annotation budgets to validate situation that shows the effectiveness of our proposal. Conclusions: We presented a class-wise confidence score to improve the acquisition function for active learning and explored unlabeled data with our proposed class-wise confidence score, which results in a large improvement over the compared methods. The experiments also showed that our proposal improved the segmentation performance for classes with a low-appearing frequency.

DOI： 10.1007/s11548-022-02773-2

A skeleton context-aware 3D fully convolutional network for abdominal artery segmentation 招待有り査読有り

Zhu, RY; Oda, M; Hayashi, Y; Kitasaka, T; Misawa, K; Fujiwara, M; Mori, K

INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY 18 巻 ( 3 ) 頁： 461 - 472 2023年3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: This paper aims to propose a deep learning-based method for abdominal artery segmentation. Blood vessel structure information is essential to diagnosis and treatment. Accurate blood vessel segmentation is critical to preoperative planning. Although deep learning-based methods perform well on large organs, segmenting small organs such as blood vessels is challenging due to complicated branching structures and positions. We propose a 3D deep learning network from a skeleton context-aware perspective to improve segmentation accuracy. In addition, we propose a novel 3D patch generation method which could strengthen the structural diversity of a training data set. Method: The proposed method segments abdominal arteries from an abdominal computed tomography (CT) volume using a 3D fully convolutional network (FCN). We add two auxiliary tasks to the network to extract the skeleton context of abdominal arteries. In addition, our skeleton-based patch generation (SBPG) method further enables the FCN to segment small arteries. SBPG generates a 3D patch from a CT volume by leveraging artery skeleton information. These methods improve the segmentation accuracies of small arteries. Results: We used 20 cases of abdominal CT volumes to evaluate the proposed method. The experimental results showed that our method outperformed previous segmentation accuracies. The averaged precision rate, recall rate, and F-measure were 95.5%, 91.0%, and 93.2%, respectively. Compared to a baseline method, our method improved 1.5% the averaged recall rate and 0.7% the averaged F-measure. Conclusions: We present a skeleton context-aware 3D FCN to segment abdominal arteries from an abdominal CT volume. In addition, we propose a 3D patch generation method. Our fully automated method segmented most of the abdominal artery regions. The method produced competitive segmentation performance compared to previous methods.

DOI： 10.1007/s11548-022-02767-0

Deep Learning-Based Seminal Vesicle and Vas Deferens Recognition in the Posterior Approach of Robot-Assisted Radical Prostatectomy. 招待有り査読有り

Takeshita N, Sakamoto S, Kitaguchi D, Takeshita N, Yajima S, Koike T, Ishikawa Y, Matsuzaki H, Mori K, Masuda H, Ichikawa T, Ito M

Urology 173 巻頁： 98 - 103 2023年3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：Urology

Objective: To develop a convolutional neural network to recognize the seminal vesicle and vas deferens (SV-VD) in the posterior approach of robot-assisted radical prostatectomy (RARP) and assess the performance of the convolutional neural network model under clinically relevant conditions. Methods: Intraoperative videos of robot-assisted radical prostatectomy performed by the posterior approach from 3 institutions were obtained between 2019 and 2020. Using SV-VD dissection videos, semantic segmentation of the seminal vesicle-vas deferens area was performed using a convolutional neural network-based approach. The dataset was split into training and test data in a 10:3 ratio. The average time required by 6 novice urologists to correctly recognize the SV-VD was compared using intraoperative videos with and without segmentation masks generated by the convolutional neural network model, which was evaluated with the test data using the Dice similarity coefficient. Training and test datasets were compared using the Mann–Whitney U-test and chi-square test. Time required to recognize the SV-VD was evaluated using the Mann–Whitney U-test. Results: From 26 patient videos, 1 040 images were created (520 SV-VD annotated images and 520 SV-VD non-displayed images). The convolutional neural network model had a Dice similarity coefficient value of 0.73 in the test data. Compared with original videos, videos with the generated segmentation mask promoted significantly faster seminal vesicle and vas deferens recognition (P < .001). Conclusion: The convolutional neural network model provides accurate recognition of the SV-VD in the posterior approach RARP, which may be helpful, especially for novice urologists.

DOI： 10.1016/j.urology.2022.12.006

Database-driven patient-specific registration error compensation method for image-guided laparoscopic surgery. 招待有り査読有り

Hayashi Y, Misawa K, Mori K

International journal of computer assisted radiology and surgery 18 巻 ( 1 ) 頁： 63 - 69 2023年1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: A surgical navigation system helps surgeons understand anatomical structures in the operative field during surgery. Patient-to-image registration, which aligns coordinate systems between the CT volume and a positional tracker, is vital for accurate surgical navigation. Although a point-based rigid registration method using fiducials on the body surface is often utilized for laparoscopic surgery navigation, precise registration is difficult due to such factors as soft tissue deformation. We propose a method that compensates a transformation matrix computed using fiducials on the body surface based on the analysis of positional information in the database. Methods: We built our database by measuring the positional information of the fiducials and the guidance targets in both the CT volume and positional tracker coordinate systems through previous surgeries. We computed two transformation matrices: using only the fiducials and using only the guidance targets in all the data in the database. We calculated the differences between the two transformation matrices in each piece of data. The compensation transformation matrix was computed by averaging these difference matrices. In this step, we selected the data from the database based on the similarity of the fiducials and the configuration of the guidance targets. Results: We evaluated our proposed method using 20 pieces of data acquired during laparoscopic gastrectomy for gastric cancer. The locations of blood vessels were used as guidance targets for computing target registration error. The mean target registration errors significantly decreased from 33.0 to 17.1 mm before and after the compensation. Conclusion: This paper described a registration error compensation method using a database for image-guided laparoscopic surgery. Since our proposed method reduced registration error without additional intraoperative measurements during surgery, it increases the accuracy of surgical navigation for laparoscopic surgery.

DOI： 10.1007/s11548-022-02804-y

Computer-Aided Size Estimation of Colorectal Polyps 招待有り査読有り

Hotta K., Itoh H., Mori Y., Misawa M., Mori K., Kudo S.e.

Techniques and Innovations in Gastrointestinal Endoscopy 25 巻 ( 2 ) 頁： 186 - 188 2023年1月

Surgical area recognition from laparoscopic images in laparoscopic gastrectomy for gastric cancer using label smoothing and uncertainty 招待有り査読有り

Hayashi Y., Misawa K., Mori K.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE 12466 巻 2023年

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：Progress in Biomedical Optics and Imaging - Proceedings of SPIE

This paper presents surgical area recognition from laparoscopic images in laparoscopic gastrectomy. Laparoscopic gastrectomy is performed as a minimally invasive procedure for removing gastric cancer. In this surgery, surgeons should cut the blood vessels around the stomach before resecting cancer. Since this type of surgery requires higher surgical skill, a surgical assistance system has been developed to enhance surgeons' abilities. Recognition of the surgical area related to the blood vessels from laparoscopic videos provides essential information in the operative field to the surgical assistance system. Therefore, we develop a method for recognizing laparoscopic images into the surgical area. The proposed method classifies the laparoscopic images into seven scenes using deep neural networks. We introduce the label smoothing in time direction to obtain a soft label. Bayesian neural networks are used to classify the laparoscopic images and estimate the uncertainty. After the classification, we modify the predictions on each laparoscopic image using the estimated uncertainty and temporal information. We evaluated the proposed method using 10,818 images from 10 videos recorded during laparoscopic gastrectomy for gastric cancer. Five-fold cross-validation was performed for the performance evaluation. The mean classification accuracy was 84.0%. The experimental results showed that the proposed method could recognize the surgical area from laparoscopic images.

DOI： 10.1117/12.2654775

AI画像解析による内視鏡外科手術手技のビデオ評価及び手術支援システムの構築招待有り査読有り Open Access

安井昭洋, 内田広夫, 森健策, 石田昇平, 出家亨一, 檜顕成, 城田千代栄, 小田昌宏, 林雄一郎

生体医工学 Annual61 巻 ( Abstract ) 頁： 127_2 - 127_2 2023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：公益社団法人日本生体医工学会

<p>【はじめに】術後成長発達する小児患者にとって、低侵襲手術は非常に重要である。しかし患者数は限られているため、しっかりとした手術を行うためにoff the job-training(OJT)が重要である。さらにOJTでの効率的な手技獲得には、手技を客観的に評価しfeed backを行うシステムが必須である。また安全で効率的な内視鏡手術を行うためには、臓器の位置関係の把握が必要であるため、術中ナビゲーションは重要な要件となる。これらの課題に対して、AIを用いた内視鏡手技評価および手術支援システムの構築に着手しており現状の成果を報告する。【方法と結果】食道閉鎖症モデルを用いた吻合手技を被験者に課し、各被験者の手技を最初に人の目で「check 表」「エラー項目」「時間」を用いて評価した。次にビデオから検出した鉗子の動きと人が判定した手技優劣の関係性をAIで学習させ、上位88％・下位95％の精度で手技優劣が自動判定可能となった。この結果を解析することで今まで必要だった50項目以上の肉眼チェックが、わずか７項目チェックするだけで手技の優劣を判断できることが明らかになった。現在食道閉鎖症の手術画像を用いて、食道・迷走神経・気管を深層学習させ、各種構造物の自動認識を進めている。【まとめ】AI画像解析により内視鏡手技の優劣をビデオで判定可能となった。この結果から新たに効率的な手技判断基準を定めることができた。術中ナビゲーションは現在精度のさらなる向上を目指している。</p>

DOI： 10.11239/jsmbe.annual61.127_2

膀胱鏡画像におけるtiny-YOLOを用いた腫瘍検出招待有り査読有り Open Access

牟田口淳, 小田昌宏, 猪口淳一, 森健策, 江藤正俊

生体医工学 Annual61 巻 ( Abstract ) 頁： 255_2 - 255_2 2023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：公益社団法人日本生体医工学会

<p>【背景】膀胱癌は経尿道手術後に再発が多い腫瘍であり、膀胱鏡での腫瘍の見落としが原因とされている。内視鏡での観察は、従来の白色光(WLI)の他に、NBIを使用するが、いずれの腫瘍検出精度は検者の技量・経験に依存するため、検査の再現性・客観性が少ないことが課題である。近年、人工知能(AI)が多くの医療分野で活用されており、AIによる検査は、客観性・再現性を持った上で、エキスパートレベルと同程度の診断能を持つ可能性があるとされている。今回、WLI/NBI膀胱鏡画像を用いて、AIによる腫瘍検出の精度を検証した。【方法】2019年から2021年まで、経尿道的膀胱腫瘍切除術(TURBT)の際に、WLI/NBIを用いて観察を行った症例の手術動画から膀胱鏡画像を作成し、腫瘍を含む画像を腫瘍画像、腫瘍を含まない画像を正常画像と定義した。腫瘍画像内の膀胱腫瘍を矩形でアノテーションを行い、テストデータ用の画像を用いてAIによる感度、特異度、陽性的中率を評価した。AIでの物体検出はtiny-YOLOを用い、腫瘍検出精度の検証を行った。【結果】WLIとNBIから、それぞれ腫瘍画像をそれぞれtiny-YOLOで学習を行い、腫瘍画像(WLI: 525枚、NBI:219枚)と正常画像(WLI:98枚、NBI:108枚)で精度検証を行った。AIによる物体検出の感度/特異度/陽性的中率は、WLIで87.8%/88.8%/97.7%、NBIで82.2%/81.4%/90.0%であった。【結論】膀胱鏡画像において、AIにより比較的良好に腫瘍検出が可能であった。更なる精度改善、リアルタイム検出への課題について、文献的考察を加え報告する。</p>

DOI： 10.11239/jsmbe.annual61.255_2

看護師の職域プレゼンティズム（心身不調）をIoTセンシングで早期検出する招待有り査読有り Open Access

山下佳子, 大山慎太郎, 鈴木輝彦, 坂本祐二, 出野義則, 山下暁士, 赤川里美, 藤井晃子, 白鳥義宗, 森健策

生体医工学 Annual61 巻 ( Abstract ) 頁： 225_1 - 225_1 2023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：公益社団法人日本生体医工学会

<p>看護職は心身不調に由来する顕在化していない労働生産性低下状態（プレゼンティズム）が他職種より多く、不規則な勤務形態や強い心身ストレスにより、離職に至る率も高いのが現状である。類似研究は腰痛に注目してウェアラブルデバイスで筋硬度測定によりプレゼンティズム発生を予測するが、性別や運動経験の有無など個人の特性を大きく反映するため予測が難しいと報告がある。そこで本研究では、プレゼンティズムの主な要因で筋骨格系疼痛の原因となる看護行動を、Internet of Things (IoT)センシングで抽出・判別することを目的とする。既設のBLE 屋内測位(AoA法）アンテナを利用した加速度センサデバイスによる看護行動データの収集と、観察者の直接観察による看護行動に対するラベルデータ（看護行動、実施場所、行動の開始・終了時刻を入力）作成を実施した。看護行動は身体プレゼンティズムリスク(腰痛・膝痛・頚肩腕痛)に分類し定義した。1分毎の行動データ（位置情報の平均値、加速度の最大・最小・平均値・標準偏差）、及び時刻を入力データとし、ラベルデータを教師データとして機械学習を行い、行動認識モデルを構築した。看護行動割合は、腰痛43.6%、膝痛3.3%、頚肩腕痛7.4%であった。IoTセンシングで個人ごとのプレゼンティズムに発展する行動をモニタリングすることにより、プレゼンティズムを予防することが可能と示唆された。</p>

DOI： 10.11239/jsmbe.annual61.225_1

医用画像とAI 招待有り

森健策

カレントテラピー 41 巻頁： 79 - 79 2023年

TriMix: A General Framework for Medical Image Segmentation from Limited Supervision 招待有り査読有り

Zheng Z., Hayashi Y., Oda M., Kitasaka T., Mori K.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 13846 LNCS 巻頁： 185 - 202 2023年

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：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

We present a general framework for medical image segmentation from limited supervision, reducing the reliance on fully and densely labeled data. Our method is simple, jointly trains triple diverse models, and adopts a mix augmentation scheme, and thus is called TriMix. TriMix imposes consistency under a more challenging perturbation, i.e., combining data augmentation and model diversity on the tri-training framework. This straightforward strategy enables TriMix to serve as a strong and general learner learning from limited supervision using different kinds of imperfect labels. We conduct extensive experiments to show TriMix’s generic purpose for semi- and weakly-supervised segmentation tasks. Compared to task-specific state-of-the-arts, TriMix achieves competitive performance and sometimes surpasses them by a large margin. The code is available at https://github.com/MoriLabNU/TriMix.

DOI： 10.1007/978-3-031-26351-4_12

Thrombosis region extraction and quantitative analysis in confocal laser scanning microscopic image sequence in in-vivo imaging 招待有り査読有り

Wu Y., Oda M., Hayashi Y., Kawamura S., Takebe T., Mori K.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE 12468 巻 2023年

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：Progress in Biomedical Optics and Imaging - Proceedings of SPIE

In this paper, we propose a scheme that includes automated extraction of thrombus regions and quantitative analysis of thrombosis in confocal laser scanning microscope (CLSM) blood flow image sequence. Making thrombosis model in animal models play an important role in the development of antithrombotic drugs and ascertaining thrombosis mechanisms. Making thrombosis model in cerebral cortex of mice is usually observed using a CLSM in the fluorescence mode. However, some small changes of thrombus regions are not easily observed in CLSM blood flow image sequences. In addition, it is not easy for researchers to quantitatively analyze the degree of thrombosis. Therefore, we propose a scheme to achieve automatic thrombosis region extraction and quantitative analysis. In which, our thrombosis region extraction method uses analysis of changing pattern of thrombosis regions in CLSM blood flow image sequence. Experimental results showed that our scheme can help biological researchers observe and analyze the changes of thrombosis in animal models and reduced the use of fluorescent thrombus markers.

DOI： 10.1117/12.2654632

Real Bronchoscopic Images-based Bronchial Nomenclature: a Preliminary Study 招待有り査読有り

Wang C., Hayashi Y., Oda M., Kitasaka T., Takabatake H., Mori M., Honma H., Natori H., Mori K.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE 12466 巻 2023年

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：Progress in Biomedical Optics and Imaging - Proceedings of SPIE

This article describes a method for bronchial nomenclature using real bronchoscopic (RB) images and pre-built knowledge base of branches. The bronchus has a complex tree-like structure, which increases the difficulty of bronchoscopy. Therefore, a bronchoscopic navigation system is used to help physicians during examination. Conventional navigation system used preoperative CT images and real bronchoscopic images to obtain the camera pose for navigation, whose accuracy is influenced by organ deformation. We propose a bronchial nomenclature method to estimate branch names for bronchoscopic navigation. This method consists of a bronchus knowledge base construction model, a camera motion estimation module, an anatomical structure tracking module, and a branch name estimation module. The knowledge base construction module is used to find the relationship of each branch. The anatomical tracking module is used to track the bronchial orifice (BO) extracted in each RB frame. The camera motion estimation module is used to estimate the camera motion between two frames. The branch name estimation module uses the pre-built bronchus knowledge base and BO tracking results to find the name of each branch. Experimental results showed that it is possible to estimate branch names using only RB images and the pre-built knowledge base of branches.

DOI： 10.1117/12.2654508

Priority attention network with Bayesian learning for fully automatic segmentation of substantia nigra from neuromelanin MRI 招待有り査読有り

Hu T., Itoh H., Oda M., Saiki S., Hattori N., Kamagata K., Aoki S., Mori K.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE 12464 巻 2023年

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：Progress in Biomedical Optics and Imaging - Proceedings of SPIE

Neuromelanin magnetic resonance imaging (NM-MRI) has been widely used in the diagnosis of Parkinson’s disease (PD) for its significantly enhanced contrast between the PD-related structure, the substantia nigra (SN) and surrounding tissues. To develop the computer-aided diagnosis (CAD) system of PD and reduce the labor burden of clinicians, precise and automatic segmentation of SN is becoming more and more desired. This paper proposes a novel network combining the priority gating attention and Bayesian learning for improving the accuracy of fully automatic SN segmentation from NM-MRI. Different from the conventional gated attention model, the proposed network uses the prior SN probability map for guiding the attention computation and reducing the potential disruptions introduced by the background. Additionally, to lower the risks of over-fitting and estimate the confidence scores for the segmentation results, Bayesian learning with Monte Carlo dropout is applied in the training and testing phases. The quantitative results showed that the proposed network acquired the averaged Dice score of 79.46% in comparison with the baseline model 77.93%.

DOI： 10.1117/12.2655112

Octree Cube Constraints in PBD Method for High Resolution Surgical Simulation 招待有り査読有り

Miyazaki R., Hayashi Y., Oda M., Mori K.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE 12464 巻 2023年

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：Progress in Biomedical Optics and Imaging - Proceedings of SPIE

This paper proposes a deformable tissue model that introduces octree lattice vertex layout and cubic constraints to the orthodox PBD (Position Based Dynamics) method. Surgical simulation is expected to provide a safe method for training in surgery, which is especially useful for preoperative education of inexperienced surgeons and/or for the case a prior attempt is required. To build a surgical simulator, it is necessary to develop organ models with deformations and interaction algorithms between surgical instruments and organ models, all of which must be performed in real time. Since existing surgical simulators focus on real-time performance, the resolution of organ models is limited. The proposed method restricts the vertex locations of the PBD method to the vertices of the octree lattice to save computation time while maintaining a high deformation resolution. To obtain appropriate results even for large deformations, three-dimensional constraints are applied to each octree cube as the constraints of the PBD method. In the simulations, we tested the overall deformation by dropping a liver model and the local deformation scene by laparoscopic clipping. As a result, we achieved deformation simulations at 26.5 fps for the model with approximately 2,672 cube elements and 20,659 vertices.

DOI： 10.1117/12.2654092

Multi-view Guidance for Self-supervised Monocular Depth Estimation on Laparoscopic Images via Spatio-Temporal Correspondence 招待有り査読有り

Li W., Hayashi Y., Oda M., Kitasaka T., Misawa K., Mori K.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 14228 LNCS 巻頁： 429 - 439 2023年

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：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

This work proposes an innovative self-supervised approach to monocular depth estimation in laparoscopic scenarios. Previous methods independently predicted depth maps ignoring spatial coherence in local regions and temporal correlation between adjacent images. The proposed approach leverages spatio-temporal coherence to address the challenges of textureless areas and homogeneous colors in such scenes. This approach utilizes a multi-view depth estimation model to guide monocular depth estimation when predicting depth maps. Moreover, the minimum reprojection error is extended to construct a cost volume for the multi-view model using adjacent images. Additionally, a 3D consistency of the point cloud back-projected from predicted depth maps is optimized for the monocular depth estimation model. To benefit from spatial coherence, deformable patch-matching is introduced to the monocular and multi-view models to smooth depth maps in local regions. Finally, a cycled prediction learning for view synthesis and relative poses is designed to exploit the temporal correlation between adjacent images fully. Experimental results show that the proposed method outperforms existing methods in both qualitative and quantitative evaluations. Our code is available at https://github.com/MoriLabNU/MGMDepthL.

DOI： 10.1007/978-3-031-43996-4_41

Masked Frequency Consistency for Domain-Adaptive Semantic Segmentation of Laparoscopic Images 招待有り査読有り

Zhao X., Hayashi Y., Oda M., Kitasaka T., Mori K.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 14220 LNCS 巻頁： 663 - 673 2023年

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：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Semantic segmentation of laparoscopic images is an important issue for intraoperative guidance in laparoscopic surgery. However, acquiring and annotating laparoscopic datasets is labor-intensive, which limits the research on this topic. In this paper, we tackle the Domain-Adaptive Semantic Segmentation (DASS) task, which aims to train a segmentation network using only computer-generated simulated images and unlabeled real images. To bridge the large domain gap between generated and real images, we propose a Masked Frequency Consistency (MFC) module that encourages the network to learn frequency-related information of the target domain as additional cues for robust recognition. Specifically, MFC randomly masks some high-frequency information of the image to improve the consistency of the network’s predictions for low-frequency images and real images. We conduct extensive experiments on existing DASS frameworks with our MFC module and show performance improvements. Our approach achieves comparable results to fully supervised learning method on the CholecSeg8K dataset without using any manual annotation. The code is available at github.com/MoriLabNU/MFC.

DOI： 10.1007/978-3-031-43907-0_63

　前のページ - 次のページ　