画像診断支援人工知能

Artificial Intelligence Enhanced Medical Robotics

Informatics, Robotics and intelligent systems

Oct. 2025 - Present
The University of Electro-Communications, Graduate School of Informatics and Engineering, Assistant Professor, Japan

Oct. 2022 - Sep. 2025
The University of Electro-Communications, 次世代研究者挑戦的研究プログラム, UEC次世代研究員, Japan

Apr. 2022 - Apr. 2022
シャープ株式会社, SSTC部門, 会社員, Japan

Oct. 2022 - Sep. 2025
The University of Electro-Communications, Graduate School of Informatics and Engineering, Department of Mechanical and Intelligent Systems Engineering, 博士後期課程, Japan

Apr. 2020 - Mar. 2022
The University of Electro-Communications, Graduate School of Informatics and Engineering, Department of Mechanical and Intelligent Systems Engineering, 博士前期課程, Japan

Sep. 2014 - Jun. 2018
上海工程技術大学, 機械工学部, 機械設計製造およびその自動化専攻, China

Mar. 2026
電気通信大学 通機会
田中栄賞, 周 家禕

Sep. 2025
IEEE EMBS Japan Chapter
3D Reconstruction of the Kidney and Lesions from 2D Ultrasound Images Using Robotic Ultrasound Diagnostic System
IEEE EMBS East and Central Japan Chapter / West Japan Chapter Young Researcher Award, Zhou Jiayi;Norihiro Koizumi;Yu Nishiyama;Peiji Chen;Kazushi Numata

Jun. 2025
IEEE UR 2025 Program Committee
A Framework for 3D Ultrasound Reconstruction Using Robotic Ultrasound Diagnostic System
Finalists for the Best Conference Paper Award of IEEE UR 2025, Jiayi Zhou;Norihiro Koizumi;Yu Nishiyama;Peiji Chen;Kazushi Numata

Jun. 2025
IEEE UR 2025 Program Committee
A Framework for 3D Ultrasound Reconstruction Using Robotic Ultrasound Diagnostic System
Finalists for the Best Application Paper Award of IEEE UR 2025, Jiayi Zhou;Norihiro Koizumi;Yu Nishiyama;Peiji Chen;Kazushi Numata

Dec. 2020
日本超音波医学会
超音波診断・治療用ロボティック・ベッドの開発
第21回日本超音波医学会奨励賞（受賞者の共同演者）, 小林賢人;佐々木雄大;江浦史生;小林賢大;渡部祐介;周 家禕;大塚 研秀;西山 悠;小泉 憲裕;沼田 和司

Development of an automated MRI-TRUS registration system focusing on the prostatic apex
A. Endo; N. Koizumi; Y. Nishiyama; P. Chen; J. Zhou; K. Yamada; G. Karakida; R. Kasagi; T. Nakamura; S. Shoji
Proc. of 40th International Congress and Exhibition on computer assisted radiology and surgery (CARS 2026), International Journal of Computer Assisted Radiology and Surgery (IJCARS), Vol.21, Suppl.1, s1-s2, Jul. 2026, Peer-reviwed
Scientific journal, English

Development of a Stepwise Real-Human-Image Similarity Transformation Method for Phantom Ultrasound Images
Miura, T.; Koizumi, N.; Yamada, K.; Nishiyama, Y.; Chen, P.; Zhou, J.; Kasagi, R.; Fujii, I.; Numata K.
Proc. of 40th International Congress and Exhibition on computer assisted radiology and surgery (CARS 2026), International Journal of Computer Assisted Radiology and Surgery (IJCARS), Vol.21, Suppl.1, s1-s2, Jul. 2026, Peer-reviwed
English

A Framework for 3D Ultrasound Reconstruction Using Robotic Ultrasound Diagnostic System
Jiayi Zhou; Norihiro Koizumi; Yu Nishiyama; Peiji Chen; Kazushi Numata
Lead, 2025 22nd International Conference on Ubiquitous Robots (UR), Jul. 2025, Peer-reviwed
International conference proceedings

3D Reconstruction of the Kidney and Lesions from 2D Ultrasound Images Using Robotic Ultrasound Diagnostic System.
Jiayi Zhou; Norihiro Koizumi; Yu Nishiyama; Peiji Chen; Kazushi Numata
Lead, Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference, 2025, 1-7, Jul. 2025, Peer-reviwed, True, This study presents an innovative approach based on the Robotic Ultrasound Diagnostic System (RUDS) for 3D reconstruction of organs and lesions from sequential 2D ultrasound slices. The RUDS comprises four main components: the Organ Tracking Robot (OTR) for multi-angle probe scanning, the Phantom Posture Robot (PPR) for optimizing probe contact with the abdomen, the Robotic Bed (RB), and the Robotic Supporting Arm (RSA). Leveraging the high flexibility and degrees of freedom in probe manipulation, the OTR controls probe rotation to capture and process continuous 2D slices, enabling precise 3D reconstruction of the kidney and associated lesions. By adjusting scanning speed and contact angles and analyzing spatial relationships within the 3D models, RUDS demonstrates superior accuracy in 3D modeling, supporting preoperative planning in High-Intensity Focused Ultrasound (HIFU) and minimizing risks to non-tumor tissue.Clinical Relevance- This enables precise 3D reconstruction of kidneys and lesions from 2D ultrasound images, supporting HIFU planning and enhancing ultrasound diagnostic accuracy.
International conference proceedings, English

Real-Time Ultrasound Image Evaluation using Deep-Learning for Automatic Ultrasound Diagnosi
Kou Koresawa; Jiayi Zhou; Riku Kasagi; Kazushi Numata; Yu Nishiyama; Norihiro Koizumi
Proc. of 20th Asian Conference on Computer-Aided Surgery (ACCAS 2024), 16-17, Sep. 2024, Peer-reviwed
International conference proceedings, English

Rib region detection for scanning path planning for fully automated robotic abdominal ultrasonography.
Koudai Okuzaki; Norihiro Koizumi; Kiyoshi Yoshinaka; Yu Nishiyama; Jiayi Zhou; Ryosuke Tsumura
International journal of computer assisted radiology and surgery, Springer Science and Business Media LLC, 19, 3, 449-457, Mar. 2024, Peer-reviwed, True, PURPOSE: Scanning path planning is an essential technology for fully automated ultrasound (US) robotics. During biliary scanning, the subcostal boundary is critical body surface landmarks for scanning path planning but are often invisible, depending on the individual. This study developed a method of estimating the rib region for scanning path planning toward fully automated robotic US systems. METHODS: We proposed a method for determining the rib region using RGB-D images and respiratory variation. We hypothesized that detecting the rib region would be possible based on changes in body surface position due to breathing. We generated a depth difference image by finding the difference between the depth image taken at the resting inspiratory position and the depth image taken at the maximum inspiratory position, which clearly shows the rib region. The boundary position of the subcostal was then determined by applying training using the YOLOv5 object detection model to this depth difference image. RESULTS: In the experiments with healthy subjects, the proposed method of rib detection using the depth difference image marked an intersection over union (IoU) of 0.951 and average confidence of 0.77. The average error between the ground truth and predicted positions was 16.5 mm in 3D space. The results were superior to rib detection using only the RGB image. CONCLUSION: The proposed depth difference imaging method, which measures respiratory variation, was able to accurately estimate the rib region without contact and physician intervention. It will be useful for planning the scan path during the biliary imaging.
Scientific journal, English

Study on method of organ section retention and tracking through deep learning in automated diagnostic and therapeutic robotics.
Takumi Fujibayashi; Norihiro Koizumi; Yu Nishiyama; Yusuke Watanabe; Jiayi Zhou; Momoko Matsuyama; Miyu Yamada; Ryosuke Tsumura; Kiyoshi Yoshinaka; Naoki Matsumoto; Hiroyuki Tsukihara; Kazushi Numata
International Journal of Computer Assisted Radiology and Surgery, Springer Science and Business Media LLC, 18, 11, 2101-2109, Nov. 2023, Peer-reviwed, In high-intensity focused ultrasound (HIFU) treatment of the kidney and liver, tracking the organs is essential because respiratory motions make continuous cauterization of the affected area difficult and may cause damage to other parts of the body. In this study, we propose a tracking system for rotational scanning, and propose and evaluate a method for estimating the angles of organs in ultrasound images.We proposed AEMA, AEMAD, and AEMAD++ as methods for estimating the angles of organs in ultrasound images, using RUDS and a phantom to acquire 90-degree images of a kidney from the long-axis image to the short-axis image as a data set. Six datasets were used, with five for preliminary preparation and one for testing, while the initial position was shifted by 2 mm in the contralateral axis direction. The test data set was evaluated by estimating the angle using each method.The accuracy and processing speed of angle estimation for AEMA, AEMAD, and AEMAD++ were 23.8% and 0.33 FPS for AEMAD, 32.0% and 0.56 FPS for AEMAD, and 29.5% and 3.20 FPS for AEMAD++, with tolerance of ± 2.5 degrees. AEMAD++ offered the best speed and accuracy.In the phantom experiment, AEMAD++ showed the effectiveness of tracking the long-axis image of the kidney in rotational scanning. In the future, we will add either the area of surrounding organs or the internal structure of the kidney as a new feature to validate the results.
Scientific journal, English

Image Search Strategy via Visual Servoing for Robotic Kidney Ultrasound Imaging.
Takumi Fujibayashi; Norihiro Koizumi; Yu Nishiyama; Jiayi Zhou; Hiroyuki Tsukihara; Kiyoshi Yoshinaka; Ryosuke Tsumura
Journal of Robotics and Mechatronics, Fuji Technology Press Ltd., 35, 5, 1281-1289, Oct. 2023, Peer-reviwed, Ultrasound (US) imaging is beneficial for kidney diagnosis; however, it involves sophisticated tasks that must be performed by physicians to obtain the target image. We propose a target-image search strategy combining visual servoing and deep learning-based image evaluation for robotic kidney US imaging. The search strategy is designed by mimicking physicians’ motion axis of the US probe. By controlling the position of the US probe along each of the motion axes while evaluating the obtained US images based on an anatomical feature extraction method via instance segmentation with YOLACT++, we are able to search for an optimal target image. The proposed approach was validated through phantom studies. The results showed that the proposed approach could find the target kidney images with error rates of 2.88±1.76 mm and 2.75±3.36°. Thus, the proposed method enables the accurate identification of the target image, which highlights its potential for application in autonomous kidney US imaging.
Scientific journal, English

Rib born detection of scan path planning for fully-automated ultrasound robotic system
K. Okuzaki; N. Koizumi; K. Yoshinaka; J. Zhou; T. Fujibayashi; R. Tsumura
International Journal of Computer Assisted Radiology and Surgery (IJCARS), 18, 1, Jun. 2023, Peer-reviwed
International conference proceedings, English

Automatic measurement system for the diameter of the inferior vena cava using deep neural network
H. Noro; N. Koizumi; Y. Nishiyama; J. Zhou T. Ishikawa; R. Tsumura; K. Yoshinaka; H. Tsukihara; N Matsumoto; R. Masuzaki; M. Ogawa; K. Numata
International Journal of Computer Assisted Radiology and Surgery (IJCARS), The Japan Society of Mechanical Engineers, 18, 1, 2P1-B20, Jun. 2023, Peer-reviwed, Compared to MRI and CT, ultrasound can visualize the internal structures of the body noninvasively and in real time, regardless of the location. However, ultrasound diagnosis has the problem that the acquisition of diagnostic images is dependent on the skill of the operator. It is extremely difficult for non-skilled operators to distinguish the inferior vena cava from the abdominal aorta, and the measurement of the inner diameter in the same cross-section is prone to variations due to operator habits, even among skilled operators. Based on the above, this paper proposes a system for measuring time-series changes in the internal diameter of the inferior vena cava.
International conference proceedings, English

Construction of organ rotation estimation system using deep learning
M. Sano; N. Koizumi; Y. Nishiyama; J. Zhou; T. Fujibayashi; M. Matsuyama; M. Yamada; T. Ishikawa; A. Katsuragi; S. Monma
18, 1, Jun. 2023, Peer-reviwed
International conference proceedings, English

A VS ultrasound diagnostic system with kidney image evaluation functions.
Jiayi Zhou; Norihiro Koizumi; Yu Nishiyama; Kiminao Kogiso; Tomohiro Ishikawa; Kento Kobayashi; Yusuke Watanabe; Takumi Fujibayashi; Miyu Yamada; Momoko Matsuyama; Hiroyuki Tsukihara; Ryosuke Tsumura; Kiyoshi Yoshinaka; Naoki Matsumoto; Masahiro Ogawa; Hideyo Miyazaki; Kazushi Numata; Hidetoshi Nagaoka; Toshiyuki Iwai; Hideyuki Iijima
Lead, International journal of computer assisted radiology and surgery, Springer Science and Business Media LLC, 18, 2, 227-246, Feb. 2023, Peer-reviwed, True, PURPOSE: An inevitable feature of ultrasound-based diagnoses is that the quality of the ultrasound images produced depends directly on the skill of the physician operating the probe. This is because physicians have to constantly adjust the probe position to obtain a cross section of the target organ, which is constantly shifting due to patient respiratory motions. Therefore, we developed an ultrasound diagnostic robot that works in cooperation with a visual servo system based on deep learning that will help alleviate the burdens imposed on physicians. METHODS: Our newly developed robotic ultrasound diagnostic system consists of three robots: an organ tracking robot (OTR), a robotic bed, and a robotic supporting arm. Additionally, we used different image processing methods (YOLOv5s and BiSeNet V2) to detect the target kidney location, as well as to evaluate the appropriateness of the obtained ultrasound images (ResNet 50). Ultimately, the image processing results are transmitted to the OTR for use as motion commands. RESULTS: In our experiments, the highest effective tracking rate (0.749) was obtained by YOLOv5s with Kalman filtering, while the effective tracking rate was improved by about 37% in comparison with cases without such filtering. Additionally, the appropriateness probability of the ultrasound images obtained during the tracking process was also the highest and most stable. The second highest tracking efficiency value (0.694) was obtained by BiSeNet V2 with Kalman filtering and was a 75% improvement over the case without such filtering. CONCLUSION: While the most efficient tracking achieved is based on the combination of YOLOv5s and Kalman filtering, the combination of BiSeNet V2 and Kalman filtering was capable of detecting the kidney center of gravity closer to the kidney's actual motion state. Furthermore, this model could also measure the cross-sectional area, maximum diameter, and other detailed information of the target kidney, which meant it is more practical for use in actual diagnoses.
Scientific journal, English

Development of a VS ultrasound diagnostic system with image evaluation functions
Jiayi Zhou; Norihiro Koizumi; Yu Nishiyama; Ryosuke Tsumura; Hiroyuki Tsukihara; Naoki Matsumoto
Lead, 2022 IEEE 11th Global Conference on Consumer Electronics (GCCE), IEEE, 699-700, 18 Oct. 2022, Peer-reviwed
International conference proceedings, English

A novel complementation method of an acoustic shadow region utilizing a convolutional neural network for ultrasound-guided therapy.
Momoko Matsuyama; Norihiro Koizumi; Akihide Otsuka; Kento Kobayashi; Shiho Yagasaki; Yusuke Watanabe; Jiayi Zhou; Yu Nishiyama; Naoki Matsumoto; Hiroyuki Tsukihara; Kazushi Numata
International Journal of Computer Assisted Radiology and Surgery, Springer Science and Business Media LLC, 17, 1, 107-119, 2022, Peer-reviwed, Noise-free ultrasound images are essential for organ monitoring during regional ultrasound-guided therapy. When the affected area is located under the ribs, however, acoustic shadow is caused by the reflection of sound from hard tissues such as bone, and the image is output with missing information in this region. Therefore, in the present study, we attempt to complement the image in the missing area.The overall flow of the complementation method to generate a shadow-free composite image is as follows. First, we constructed a binary classification method for the presence or absence of acoustic shadow on a phantom kidney based on a convolutional neural network. Second, we created a composite shadow-free image by searching for a suitable image from a time-series database and superimposing the corresponding area without shadow onto the missing area of the target image. In addition, we constructed and verified an automatic kidney mask generation method utilizing U-Net.The complementation accuracy for kidney tracking could be enhanced by template matching. Zero-mean normalized cross-correlation (ZNCC) values after complementation were higher than that of before complementation under four different data generation conditions: (i) changing the position of the bed of the robotic ultrasound diagnostic system in the translational direction, (ii) changing the probe angle in the translational direction, (iii) with the addition of rotational motion of the probe to condition (ii). Although there was large variation in the shape of the kidney contour in condition (iii), the proposed method improved the ZNCC value from 0.5437 to 0.5807.The effectiveness of the proposed method was demonstrated in phantom experiments. Verification of its effectiveness in real organs is necessary in future study.
Scientific journal, English

深層学習を用いた下大静脈径自動計測システムの構築
野呂 悠紀; 小泉 憲裕; 石川 智大; Zhou Jiayi; 月原 弘之
(一社)日本コンピュータ外科学会, Nov. 2023, 日本コンピュータ外科学会誌, 25, 3, 177-177, Japanese, 1344-9486, 1884-5770, 2024082293

超音波画像による臓器の三次元モデル再構成の手法検討
石川 智大; 小泉 憲裕; Zhou Jiayi; 西山 悠
(一社)日本コンピュータ外科学会, Nov. 2023, 日本コンピュータ外科学会誌, 25, 3, 210-210, Japanese, 1344-9486, 1884-5770, 2024082319

超音波診断ロボットを用いた移動量推定システムの構築
佐野 元康; 小泉 憲裕; 西山 悠; Jiayi Zhou; 石川 智大; 桂木 嵐
(一社)日本コンピュータ外科学会, Nov. 2023, 日本コンピュータ外科学会誌, 25, 3, 241-241, Japanese, 1344-9486, 1884-5770, 2024082344

超音波画像処理と位置情報を融合した新技術展開 深層学習に基づくロボティック超音波診断支援システムの開発
Zhou Jiayi; 小泉 憲裕; 西山 悠; 津村 遼介; 葭仲 潔; 松本 直樹; 小川 眞広; 沼田 和司
(公社)日本超音波医学会, Apr. 2023, 超音波医学, 50, Suppl., S507-S507, Japanese, 1346-1176, 1881-9311, 2023261399

Development of an out-of-plane motion tracking system for organs using an ultrasound robot
KORESAWA Kou; KOIZUMI Norihiro; SANO Motoyasu; NISHIYAMA Yu; KATSURAGI Arashi; ZHOU Jiayi; TSUKIHARA Hiroyuki; ISHIKAWA Tomohiro; TSUMURA Ryosuke
Unlike MRI and CT, Ultrasonography is invasive and safety. On the other hand, ultrasound probe manipulation has high degrees of freedom, and the quality of the acquired ultrasound images are highly dependent on the habits and skills of the probe operator. In addition, it is difficult to maintain the same cross-section of organs due to organ movement caused by respiration. Especially, out-of-plane motions lead to dramatic image changes and becomes difficult to track the target organs. Based on the above, the objective of this study is to establish the construction methodology of the organ tracking system especially for the out-of-plane motion of the ultrasound image of the organ.In this report, the kidney was targeted because of its convenience of being able to fit the entire image within the ultrasound image plane., The Japan Society of Mechanical Engineers, 2023, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2023, 2P1-B19, Japanese, 2424-3124

Rib detection for Initial positioning of robotic ultrasonography
NISHIYAMA Yu; OKUZAKI Kodai; ZHOU Jiayi; KOIZUMI Norihiro; SANO Motoyasu; ISHIKAWA Tomohiro; YOSHINAKA Kiyoshi; TSUMURA Ryosuke
Scanning path planning is an essential technology for fully automated ultrasound (US) robotics. During abdominal US scanning, the ribs are critical body surface landmarks for scanning path planning. We proposed a method for determining the rib region using RGB-D images and respiratory variation. We generated a depth difference image by finding the difference between the depth image taken at the resting inspiratory position and the depth image taken at the maximum inspiratory position, which clearly shows the rib position. The rib position was then determined by yolov5 object detection model to this depth difference image. In the experiment conducted with nine subjects, the proposed method of rib detection using depth difference images marked an IoU of 0.951 and average confidence of 0.77. The average error between the ground truth and predicted positions was 15 pixels ≃ 7.5 mm. The results were superior to rib detection using only the RGB image., The Japan Society of Mechanical Engineers, 2023, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2023, 2P2-B15, Japanese, 2424-3124

A Research on Organ Tracking and Ultrasound Image Appropriateness Evaluation Based on Deep Learning
Jiayi ZHOU; ISHIKAWA Tomohiro; KORESAWA Kou; MATSUMOTO Naoki; KOIZUMI Norihiro; TSUKIHARA Hiroyuki; NISHIYAMA Yu
Ultrasound-based diagnoses' quality depends on the physician's skill in handling the probe, which can be affected by patient respiratory motions. To address this, a diagnostic robot with a deep learning-based visual servo system was developed. It uses two image processing methods, BiSeNet V2 for target kidney location detection and ResNet 50 for evaluating ultrasound image appropriateness, aiming to alleviate burdens on physicians., The Japan Society of Mechanical Engineers, 2023, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2023, 2P2-B17, Japanese, 2424-3124

Construction of a Three-dimensional Model of an Organ Using Visual SLAM and Deep Learning
NISHIYAMA Yu; KOIZUMI Norihiro; ZHOU Jiayi; ISHIKAWA Tomohiro
Ultrasonography is less invasive and safer than MRI or CT. However, image acquisition is dependent on the skill of the probe operator, and untrained inspectors and patients undergoing diagnosis have difficulty in understanding the three-dimensional structure of organs. For the above reasons, we propose a system to construct three-dimensional organ models from two-dimensional ultrasound images. The construction of a three-dimensional model requires ultrasound images and three-dimensional position information of the ultrasound probe. We attached a camera to the ultrasound probe and used Visual SLAM to estimate the position of the probe. The obtained ultrasound images are segmented using deep learning, and a 3D model of the organ is constructed based on the position information of the ultrasound probe. In this study, the right kidney of the phantom is targeted, and the results show that the position estimation and segmentation are highly accurate. Based on these results, a model of the right kidney was constructed., The Japan Society of Mechanical Engineers, 2023, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2023, 2P2-B18, Japanese, 2424-3124

自動診断治療ロボットにおける深層学習を援用したAEMAによる同一断面追従手法の研究
藤林 巧; 渡部 祐介; Zhou Jiayi; 松山 桃子; 山田 望結; 沼田 和司; 月原 弘之; 松本 直樹; 西山 悠; 小泉 憲裕
(公社)日本超音波医学会, Apr. 2022, 超音波医学, 49, Suppl., S818-S818, Japanese, 1346-1176, 1881-9311, 2022331708

Detection of Organs and Identification of Missing Diagnostic Images Using Deep Learning for Assessment of Diagnostic Image Adequacy
KATSURAGI Arashi; FUJIBAYASHI Takumi; NISHIYAMA Yuu; WATANABE Yusuke; KOIZUMI Norihiro; ZHOU Jiayi; MATSUYAMA Momoko; YAMADA Miyu; YOSHINAKA Kiyoshi; TSUMURA Ryosuke; TSUKIHARA Hiroyuki; NUMATA Kazushi; MATSUMOTO Naoki; MASUZAKI Ryouta; OGAWA Masahiro
The purpose of this study was to evaluate the appropriateness of diagnostic images for automated ultrasound operations. Therefore, two experiments were conducted. The first is to detect the target organ using deep learning. The second is to identify missing parts in the diagnostic images. In the first experiment of organ detection, the IoU and Dice coefficient were 0.947 and 0.972, respectively, indicating high accuracy.In the second experiment to identify the missing parts of the image, the percentage of correct answers for the missing parts on the right side of m was 75.3%, while the percentage of correct answers for the missing parts on the left side was 99.1%., The Japan Society of Mechanical Engineers, 2022, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2022, 1P1-M12, Japanese, 2424-3124

Proposal of an Automatic Probe Manipulation Model Considering Acoustic Shadow in Ultrasound Diagnosis
ISHIKAWA Tomohiro; YAMADA Miyu; TSUMURA Ryosuke; MATSUMOTO Naoki; YOSHINAKA Kiyoshi; TSUKIHARA Hiroyuki; NUMATA Kazushi; FUJIBAYASHI Takumi; MATSUYAMA Momoko; NISHIYAMA Yu; WATANABE Yusuke; ZHOU Jiayi; KOIZUMI Norihiro; YAGASAKI Shiho
In ultrasound therapy, a clear ultrasound image is necessary to determine the exact irradiation position. However, there is a concern that the accuracy of irradiation may be degraded due to the black noise caused by the reflection of sound waves on hard tissues such as ribs and stones. In this study, we aim to automate ultrasound probe manipulation to support monitoring of ultrasound diagnosis. The acoustic shadow and the target organ in the ultrasound image are detected by deep learning, and the control model avoids overlapping imaging in real time based on the overlapping area information of the two. This makes it possible to monitor the treatment target without any acoustic shadows., The Japan Society of Mechanical Engineers, 2022, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2022, 1P1-M11, Japanese, 2424-3124

Development of a system to imitation the motion trajectory of a probe using Visual SLAM for an automated ultrasound diagnostic robot
Monma Sho; Nishiyama Yu; Koizumi Norihiro; Jiayi Zhou; Ishikawa Tomohiro; Matsuyama Momoko; Watanabe Yusuke; Fujibayashi Takumi; Tsumura Ryosuke; Yamada Miyu; Yoshinaka Kiyoshi; Ogawa Masahiro; Matsumoto Naoki; Tsukihara Hiroyuki; Numata Kazushi
The purpose of this study is to estimate the motion trajectory of an ultrasound probe using Visual SLAM technology，and to reproduce a doctor's probe scanning using a bed-type robotic ultrasound diagnosis system (RUDS)． We investigated a method to mimic the motion trajectory of the RUDS probe by using SLAM technology．The proposed method was able to perform the imitation motion，but the error became larger as the distance from the origin increased．, The Japan Society of Mechanical Engineers, 2022, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2022, 1P1-M02, Japanese, 2424-3124

Construction of a 3D Model of Kidney Using Visual SLAM and Deep Learning with Robotic Ultrasound
KOIZUMI Norihiro; ZHOU Jiayi; WATANABE Yusuke; ISHIKAWA Tomohiro; NISHIYAMA Yu; FUJIBAYASHI Takumi; YAMADA Miyu; MATSUYAMA Momoko; TSUKIHARA Hiroyuki; NUMATA Kazushi; YOSHINAKA Kiyoshi; TSUMURA Ryosuke
Ultrasonography is less invasive and safer than MRI or CT. However, image acquisition is dependent on the skill of the probe operator, and it is difficult for an untrained examiner to understand the three-dimensional structure of the organ. For the above reason, we use a robot to acquire ultrasound images and estimate the position of the probe using Visual SLAM. The obtained images are segmented and combined with position information to construct a three-dimensional model of the organ. In this study, the right kidney of the phantom was used as the target, and the results of position estimation and segmentation accuracy were high. The model of the right kidney was constructed based on these results., The Japan Society of Mechanical Engineers, 2022, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2022, 1P1-L04, Japanese, 2424-3124

超音波医用画像における音響陰影による欠損領域の補完に関する研究
松山 桃子; 小泉 憲裕; 西山 悠; 矢ヶ崎 詞穂; 小林 賢人; 渡部 祐介; Zhou Jiayi; 松本 直樹; 月原 弘之; 沼田 和司
(公社)日本超音波医学会, Apr. 2021, 超音波医学, 48, Suppl., S621-S621, Japanese, 1346-1176, 1881-9311, 2021348391

超音波診断ロボットにおける深層学習を援用した臓器の同一断面維持・追従に関する研究
藤林 巧; 小林 賢人; Zhou Jiayi; 渡部 祐介; 松山 桃子; 山田 望結; 小泉 憲裕; 月原 弘之; 沼田 和司; 飯島 秀幸
(公社)日本超音波医学会, Apr. 2021, 超音波医学, 48, Suppl., S622-S622, Japanese, 1346-1176, 1881-9311, 2021348392

超音波診断ロボットを用いた運動状態推定及び適正画像判定システムの構築
Zhou Jiayi; 小林 賢人; 渡部 祐介; 藤林 巧; 山田 望結; 松山 桃子; 月原 弘之; 松本 直樹; 西山 悠; 小泉 憲裕
(公社)日本超音波医学会, Apr. 2021, 超音波医学, 48, Suppl., S651-S651, Japanese, 1346-1176, 1881-9311, 2021348437

超音波診断ロボットのための患部抽出追従手法に関する研究
山田 望結; Zhou Jiayi; 小林 賢人; 渡部 祐介; 藤林 巧; 松山 桃子; 小泉 憲裕; 月原 弘之; 西山 悠; 沼田 和司
(公社)日本超音波医学会, Apr. 2021, 超音波医学, 48, Suppl., S651-S651, Japanese, 1346-1176, 1881-9311, 2021348438

Development of vertical axis rotation mechanism of bed type ultrasonic robot and estimation method of posture of moving organ
SASAKI Yudai; KOBAYASHI Takahiro; WATANABE Yusuke; NISHIYAMA Yu; KOBAYASHI Kento; NUMATA Kazushi; IIJIMA Hideyuki; KOIZUMI Norihiro; ZHOU Kai; IWAI Toshiyuki; TSUKIHARA Hiroyuki; NAGAOKA Hidetoshi
In this research, we developed a mechanism that has vertical axis rotation in the tip of a bed-type ultrasound robot that can follow three-dimensional movement of abdominal organs with higher accuracy, which is impossible with a conventional bed-type ultrasound robot. As a preparation stage for position and orientation tracking, a new method for estimating the organ angle from acquired ultrasound images alone is newly proposed. From the experimental results, an appropriate angle could be estimated from the acquired data. As future work, it is required to perform an experiment to follow the organ in the vertical axis rotation with the developed mechanism and the proposed algorithm., The Japan Society of Mechanical Engineers, 2020, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2020, 2A1-E12, Japanese, 2424-3124, 130007943854

A contact state adjustment method to enhance organ motion compensation performance for a bed-type ultrasound diagnostic and therapeutic robot
WATANABE Yusuke; TSUKIHARA Hiroyuki; IIJIMA Hideyuki; Iwai Toshiyuki; Nagaoka Hidetoshi; NISHIYAMA Yu; NUMATA Kazushi; SASAKI Yudai; KOIZUMI Norihiro; KOBAYASHI Kento; ZHOU Jiayi; KOBAYASHI Takahiro
It is difficult to perform proper ultrasound diagnosis when the respiratory organs are moving, so it is necessary for the patient to stop breathing. However, stopping breathing puts a burden on the patient, so it is necessary to acquire a still image while breathing, that is, even when the organ is moving. We developed a bed-type ultrasound robot and used template matching for tracking and load cell for measuring the contact force. The subject of the tracking experiment was a phantom that imitated the abdominal organ, and the contact force was measured using the phantom. Tracking accuracy was higher at a bed speed of 1.4 cm/s on the back when the contact was sufficient on the abdomen. As for the contact force, a value of 3.0 to 4.9 N was measured on the back., The Japan Society of Mechanical Engineers, 2020, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2020, 2A1-E03, Japanese, 2424-3124, 130007943902

Study on automatic acquisition of diagnostic images by ultrasound diagnostic robot
Watanabe Yusuke; Sasaki Yudai; Kobayashi Takahiro; Kobayashi Kento; Tsukihara Hiroyuki; Iijima Hideyuki; Numata Kazushi; Iwai Toshiyuki; Nishiyama Yu; Koizumi Norihiro; Nagaoka Hidetoshi; Jiayi Zhou
The purpose of this study is to develop a novel algorithm that can automatically acquire an ultrasound image using the newly developed bed-type ultrasound diagnostic robot. We confirmed the effectiveness of our novel proposed method by comparing tracking accuracy with the conventional template matching method and simulated a basic algorithm for controlling the robot. As a result, it was confirmed that the accuracy for detecting the target is dramatically enhanced so as to obtain the target ultrasound image successfully., The Japan Society of Mechanical Engineers, 2020, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2020, 2A1-E14, Japanese, 2424-3124, 130007943853

Estimation of Organ State utilizing Deep Learning Tiny-YOLOv3 and Fixed Lag Smoothing Motion Filter
KOBAYASHI Kento; ZHOU Jiayi; SASAKI Yudai; EURA Fumio; NISHIYAMA Yu; KOIZUMI Norihiro; TSUKIHARA Hiroyuki; MATSUMOTO Naoki
In this research, we proposed a robotic motion control framework that can detect kidney in real-time by utilizing deep learning, and evaluate the accuracy of automatically acquiring and maintaining ultrasound diagnostic images of kidney. Furthermore, we performed object detection experiments using a model that was generated by a kidney phantom, estimated the state of renal phantom motion.

The novelty of our method is the framework of the combination of the deep learning tiny-YOLOv3 model and filtering considering the influence of speckle noise in an ultrasound image. In our method, the filtering is determined to incorporate the center position of where the object is detected, and two types of filters are adopted., The Japan Society of Mechanical Engineers, 2020, The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec), 2020, 2A1-E02, Japanese, 2424-3124, 130007943896

AI画像解析に基づく高精度自動超音波検査支援技術の開発補助事業
公益財団法人JKA, JKA補助事業 若手研究, 電気通信大学, Principal investigator, 2026M-295
Apr. 2026 - Mar. 2027

測定装置、測定システム、測定方法及びプログラム
Patent right, 小泉憲裕, 野呂悠紀, 石川智大, 月原弘之, 周 家禕, 西山悠, 松本直樹, 特願2023-073424, Date applied: 27 Apr. 2023, 国立大学法人電気通信大学, 学校法人日本大学

画像評価装置及びプログラム
Patent right, 小泉憲裕, 西山 悠, 小林賢人, 周 家禕, 渡部祐介, 藤林 巧, 松山桃子, 山田望結, 五十嵐立樹, 草原健太, 矢ケ崎詞穂, 沼田和司, 月原弘之, 宮嵜英世, 松本直樹, 小川眞広, 特願2022-029940, Date applied: 28 Feb. 2022, 国立大学法人電気通信大学