研究者詳細 - 神谷　亨

論文 - 神谷　亨

分割表示757 件中 281 - 300 件目／全件表示 >>

YOROを用いた海中生物のロバスト自律追跡

陸、金

動的画像処理実用化ワークショップ 65 - 66 2019年03月

Clinical Applications of Multidisciplunary Computational Anatomy to Diagnosis

Kido, Hashimoto, Hirano, Mabu, Kim, Kumura, Noriki, Inai, Tachibana

The 5th International Symposium on Mutidisciplinary Computational Anatomy 202 - 212 2019年03月

Recognition of surrounding environment from electric wheelchair videos based on modified YOLOv2 査読有り国際誌

Sakai Y., Lu H., Tan J., Kim H.

Future Generation Computer Systems 92 157 - 161 2019年03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）

© 2018 Elsevier B.V. Currently, the aging population is growing in Japan, and the needs for the utilization of welfare equipment are consequently increasing. The electric wheelchair, which is a convenient transportation tool, has rapidly become popular. However, many accidents have occurred when using electric wheelchairs, and the dangers of driving have been noted. Therefore, there is a need to improve accident factors, reduce accidents and improve the convenience of electric wheelchairs by using automation. Environmental recognition is the key technology for developing autonomous electric wheelchairs. Environmental recognition includes self-position estimation, the recognition of sidewalks, crosswalks and traffic lights, and moving object predictions. To solve these problems, this paper develops a system for detecting sidewalks, crosswalks and traffic lights. We develop the object recognition methods using a modified YOLOv2, which is an object detection algorithm that applies convolutional neural networks (CNNs). We detect the object through YOLOv2 and perform processing steps, such as unnecessary bounding box deletion and interpolation. The experimental results demonstrate that the average AUC of the detection rate is 0.587.

DOI： 10.1016/j.future.2018.09.068

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85054855462&origin=inward

CNNを用いた指骨CR画像からの骨粗しょう症の自動識別査読有り

畠野、村上、植村、陸、金、青木

Medical Imaging Technology 37 ( 2 ) 107 - 115 2019年03月

Dust removal from high turbid underwater images using convolutional neural networks 査読有り国際誌

Li Y., Zhang Y., Xu X., He L., Serikawa S., Kim H.

Optics and Laser Technology 110 2 - 6 2019年02月

記述言語：英語掲載種別：研究論文（学術雑誌）

© 2017 Elsevier Ltd In recent years, underwater image processing has been a focus of many studies. Most underwater image processing technologies are focused on descattering, absorption correction and reflection correction. For a deep sea mining machine, dust seriously affects visual acuity. To correct for this, a two-part dust removal approach is proposed. Underwater red-green minimum channel prior descattering is used to remove fine dust in the first stage. However, the impact of dust streaks on images is always undesirable. Consequently, we propose a further deep convolutional neural-network-based dust removal method. The experimental results show that the proposed method performs better in removing haze-like scatters and dust-like scatters.

DOI： 10.1016/j.optlastec.2017.09.017

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85029658908&origin=inward

Statistical shape model building method using surface registration and model prototype 査読有り国際誌

Li G., Wu J., Xiao Z., Kim H., Ogunbona P.

Optics and Laser Technology 110 234 - 238 2019年02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）

© 2017 Elsevier Ltd Automatic segmentation of organs from medical images is indispensable for the computer-assisted medical applications. Statistical Shape Models (SSMs) based scheme has been developed as an accurate and robust approach for extraction of anatomical structures, in which a crucial step is the need to place the sampled points (landmarks) with well corresponding across the whole training set. On the one hand, the correspondence of landmarks is related the quality of shape model. On the other hand, in clinical application some key positions of landmarks should be specified by physicians referring to the anatomic structure. In this paper, we develop an interactive method to build SSM that the landmark distribution can be modified manually without influencing the model quality. We extend an existing remeshing method to produce a model prototype in advance and surface features driven registration to insure the universal optimization of correspondence. The key landmarks are fixed during the prototype generation. We experimented and evaluated the proposed SSM method for lung regions, the deformations of which are considerable large.

DOI： 10.1016/j.optlastec.2017.09.018

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85030029527&origin=inward

Retinal vessel grading in fundoscopic images using GANs 査読有り

Guangxu Li, Ruan, Wan, Yan, Xiao, Kim, Deheng Li:

International Symposium on Artificial Life and Robotics 418 - 422 2019年01月

Segmentation of Spine Region in CT Images Using 3D Edge Detection and 3D Region Growing Technique 査読有り

G. Fu, H. Lu, G. Li, H. Kim, S. Murakami, M. Ueno, T. Terasawa, X. Zhu, T. Aoki

International Symposium on Artificial Life and Robotics 95 - 98 2019年01月

Ensemble 3D residual network (E3D-ResNet) for reduction of false-positive polyp detections in CT colonography 査読有り国際誌

Uemura T., Näppi J., Lu H., Kim H., Tachibana R., Hironaka T., Yoshida H.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE 10950 2019年01月

担当区分：責任著者記述言語：英語掲載種別：研究論文（国際会議プロシーディングス）

© 2019 SPIE. We developed a novel ensemble three-dimensional residual network (E3D-ResNet) for the reduction of false positives (FPs) in computer-aided detection (CADe) of polyps on CT colonography (CTC). To capture the volumetric multiscale information of CTC images, each polyp candidate was represented with three different sizes of volumes of interest (VOIs), which were enlarged to a common size and were individually subjected to three 3D-ResNets. These 3D-ResNets were trained to calculate three polyp-likelihood probabilities, p1, p2 and p3, corresponding to each input VOI. The final polyp likelihood, p, was obtained as the maximum of p1, p2 and p3. We compared the classification performance of the E3D-ResNet with that of a non-ensemble 3D-ResNet, ensemble 2D-ResNet, and ensemble of 2D-and 3D-convolutional neural network (CNN) models. All models were trained and evaluated with 21,021 VOIs of polyps and 19,557 VOIs of FPs that were sampled with data augmentation from the CADe detections on the CTC data of 20 patients. We evaluated the classification performance of the models with receiver operating characteristics (ROC) analysis using cross-validation, where the area under the ROC curve (AUC) was used as the figure of merit. Preliminary results showed that AUC value (0.98) of the E3D-ResNet was significantly higher than that of the reference models (P < 0.001), indicating that the E3D-ResNet has the potential of substantially reducing the FPs in CADe of polyps on CTC.

DOI： 10.1117/12.2512173

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85068173052&origin=inward

Low-Light Underwater Image Enhancement for Deep-Sea Tripod 査読有り国際誌

Li Y., Li J., Li Y., Kim H., Serikawa S.

IEEE Access 7 44080 - 44086 2019年01月

記述言語：英語掲載種別：研究論文（学術雑誌）

© 2013 IEEE. To monitor the sedimentary process and morphological evolution in the South China Sea, free-ascending deep-sea tripod (FDT) has been developed. This FDT was equipped with a deep-sea camera and landed on the sea floor at a depth of 2100 m. Although the FDT was equipped with an artificial light, the battery capacity limited the duration and intensity of light. Therefore, enhancing such low-illumination images to obtain clear visual effects is an important advancement for analyzing the geological evolution process. In this paper, an adaptive bright-color channel-based low-light underwater image-enhancement method and a denoising method are proposed to enhance such images and remove noise and artifacts. The experimental results demonstrated that the proposed method outperformed state-of-the-art methods.

DOI： 10.1109/ACCESS.2019.2897691

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85064689140&origin=inward

CNNを用いた指骨CR画像からの骨粗しょう症の自動識別査読有り

畠野和裕, 村上誠一, 植村知規, 陸慧敏, 金亨燮, 青木隆敏

Ｍｅｄｉｃａｌ　Ｉｍａｇｉｎｇ　Ｔｅｃｈｎｏｌｏｇｙ ( 日本医用画像工学会 ) 37 ( 2 ) 107 - 115 2019年01月

担当区分：最終著者記述言語：日本語掲載種別：研究論文（学術雑誌）

<p>骨のおもな疾患として，骨粗しょう症が挙げられる．骨粗しょう症に対する画像診断は有効であるが，医師の負担増加や経験差による診断結果のばらつき，病変部の未検出等が懸念されている．そこで本稿では，指骨computed radiography（CR）画像から骨粗しょう症の識別を行い，医師に提示するための診断支援手法を提案する．提案手法では，畳み込みニューラルネットワークの一種である，Residual Network （ResNet）を用いた識別器を構築し，骨粗しょう症有無の識別を行う．ResNetへの入力画像には，CR画像から生成した画像を用いる．本稿では，3種類の入力画像を提案し，各画像で学習および，識別の評価を行う．実験では，101症例に対し提案手法を適用し，receiver operating characteristics（ROC）曲線上のarea under the curve（AUC）値を用いて評価したところ，最大で0.931という結果を得た．</p>

DOI： 10.11409/mit.37.107

CiNii Article

CiNii Research

その他リンク： https://ci.nii.ac.jp/naid/130007633096

畳み込みニューラルネットワークを用いた指骨CR画像からの骨粗しょう症の識別査読有り

畠野和裕, 村上誠一, 植村知規, 陸慧敏, 金亨燮, 青木隆敏

医用画像情報学会雑誌 ( 医用画像情報学会 ) 36 ( 2 ) 72 - 76 2019年01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）

<p>Osteoporosis is known as one of the main diseases of bone. Although image diagnosis for osteoporosis is effective, there are concerns about increased burden of radiologists associated with diagnostic imaging, uneven diagnostic results due to experience difference, and undetected lesions. Therefore, in this study, we propose a diagnosis supporting method for classifying osteoporosis from phalanges computed radiography images and presenting classification results to physicians. In the proposed method, we construct classifiers using convolution neural network and classify normal cases and abnormal cases about osteoporosis. In our experiments, two kinds of CNN models were constructed using input images generated from 101 cases of CR images and evaluated using Area Under the Curve（AUC）value on Receiver Operating Characteristics（ROC）curve. Finaly, AUC of 0.995 was obtained.</p>

DOI： 10.11318/mii.36.72

CiNii Article

CiNii Research

その他リンク： https://ci.nii.ac.jp/naid/130007668726

Automatic extraction of abnormalities on temporal CT subtraction images using sparse coding and 3D-CNn 査読有り国際誌

Koizumi Y., Miyake N., Lu H., Kim H., Murakami S., Aoki T., Kido S.

International Conference on Control, Automation and Systems 2018-October 1468 - 1471 2018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（国際会議プロシーディングス）

© ICROS. In recent years, the proportion of deaths from cancer tends to increase in Japan, especially the number of deaths from lung cancer is increasing. CT device is effective for early detection of lung cancer. However, there is concern that an increase in burden on doctors will be caused by high performance of CT improving. Therefore, by presenting the “second opinion” by the CAD system, it reduces the burden on the doctor. In this paper, we develop a CAD system for automatic detection of lesion candidate regions such as lung nodules or ground glass opacity (GGO) from 3D CT images. Our proposed method consists of three steps. In the first step, lesion candidate regions are extracted using temporal subtraction technique. In the second step, the image is reconstructed by sparse coding for the extracted region. In the final step, 3D Convolutional Neural Network (3D-CNN) identification using reconstructed images is performed. We applied our method to 51 cases and True Positive rate (TP) of 79.81 % and False Positive rate (FP) of 37.65 % are obtained.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060466308&origin=inward

ROI-based fully automated liver registration in multi-phase CT images 査読有り国際誌

Saito K., Lu H., Kim H., Kido S., Tanabe M.

International Conference on Control, Automation and Systems 2018-October 645 - 649 2018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（国際会議プロシーディングス）

© ICROS. In this paper, we propose a registration method for fully automated liver tumor detection. Multiple phases CT is used for the detection of the liver tumor because multiple phase CT can give different characteristic features of lesions for each time phases. Registration accuracy is important when obtaining image features from multiple time phases. However, since each time phases have different image density characteristics, therefore registration of multi-phase CT is a challenging task. In this paper, we propose a robust initial alignment method independent of changing image density features in each time phase, and deformable registration method with region of interests (ROI) as liver region extracted by U-Net. Our proposed method is evaluated on 15 patient image sets. This method is applied to the early arterial phase and the equilibrium phase to registries. Experimental results show that segmentation of early arterial phase is 83% and registration is 93% accuracy.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060464031&origin=inward

Registration of phalange region from CR images based on genetic algorithm 査読有り国際誌

Kawagoe K., Murakami S., Lu H., Tan J., Kim H., Aoki T.

International Conference on Control, Automation and Systems 2018-October 1464 - 1467 2018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（国際会議プロシーディングス）

© ICROS. In Japan, the number of patients with osteoporosis and rheumatoid arthritis is increasing. Image diagnosis using CR images is effective for osteoporosis and rheumatoid arthritis. Development of a CAD system is important for reducing burdens on doctors. In this paper, we propose an automatic registration algorithm in the CAD system. In the proposed method, the genetic algorithm is used to register bone regions between identical parts of the same subject with different time series. In the experiment, the proposed method is applied to 176 bone area, and 98.14 % of TPR, 1.85 % of FPR are obtained respectively. Even when the area difference is used as the fitness of the genetic algorithm, it has cross-correlation and positioning accuracy equivalent to mutual information.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060451078&origin=inward

Extraction of median plane from facial 3D point cloud based on symmetry analysis using ICP algorithm 査読有り国際誌

Yamada S., Lu H., Tan J., Kim H., Kimura N., Okawachi T., Nozoe E., Nakamura N.

International Conference on Control, Automation and Systems 2018-October 1347 - 1350 2018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（国際会議プロシーディングス）

© ICROS. Cleft lip is a kind of congenital facial morphological abnormality. In the clinical field of cleft lip, it is necessary to analyze symmetric shape. However, there is no method to analyze the cleft lip technique based on symmetrical viewpoints. On the other hand, in our previous method to find a symmetric axis using a 2D image, since the middle line is extracted only from the front view of the face moire image. There was a problem that low accuracy was obtained by slight rotation of the face and it was not possible to consider 3D information. In this paper, we propose a method to extract the median plane of the face by analyzing based on bilateral symmetry by using 3D point cloud on the face of front. By extracting the median plane, we believe that not only surgical assistance of doctor be possible but also become a clue to development of simulation software which is the end goal.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060498535&origin=inward

Enhancement of bone metastasis from CT images based on salient region feature registration 査読有り国際誌

Sato S., Lu H., Kim H., Murakami S., Ueno M., Terasawa T., Aoki T.

International Conference on Control, Automation and Systems 2018-October 1329 - 1332 2018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（国際会議プロシーディングス）

© ICROS. In recent years, the development of the computer-aided diagnosis (CAD) systems to support radiologist is attracting attention in medical research field. One of them is temporal subtraction technique. It is a technique to generate images emphasizing temporal changes in lesions by performing a differential operation between current and previous image of the same subject. In this paper, we propose an image registration method for image registration of current and previous image, to generate temporal subtraction images from CT images and enhanced bone metastasis region. The proposed registration method is composed into three main steps: i) automatic segmentation of the region of interest (ROI) using position information of the spine based on biology, ii) use global image matching to select pairs from previous and current image, and iii) final image matching based on salient region feature. We perform registration technique on synthetic data and confirm usefulness of the proposed method. Furthermore, radiologist conduct comparative experiments without and with temporal subtraction images created by proposed method. As a result, they show high reading performance by using temporal subtraction images.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060475609&origin=inward

Detection of phalange region based on U-Net 査読有り国際誌

Hatano K., Murakami S., Lu H., Tan J., Kim H., Aoki T.

International Conference on Control, Automation and Systems 2018-October 1338 - 1342 2018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（国際会議プロシーディングス）

© ICROS. Osteoporosis is one of the famous bone diseases. It is a major cause of deteriorating the quality of life, and early detection and early treatment are becoming socially important. Visual screening using Computed Radiography (CR) images is effective for diagnosis of osteoporosis, but there are problems of increasing the burden on doctors, variation in diagnostic results due to differences in experiences of doctors, and undetected lesions. In order to solve this problem, we are working on a computer-aided diagnosis (CAD) system for osteoporosis. In this paper, we propose segmentation methods of the phalange region from the phalangeal CR images as a preprocessing of classification of osteoporosis. In the proposed method, we construct a segmentation model using U-Net, which is a type of deep convolution neural network (DCNN). The proposed method was applied to input images generated from CR images of 101 patients with both hands, and evaluated using the Intersection over Union (IoU) values. The result was 0.914 in IoU.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060468143&origin=inward

Detection of grasping position from video images based on SSD 査読有り国際誌

Kitayama T., Lu H., Li Y., Kim H.

International Conference on Control, Automation and Systems 2018-October 1472 - 1475 2018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（国際会議プロシーディングス）

© ICROS. Recently, consistent container transportation of roads and ships is mainstream of international freight transport. Because of various factors, automation of cargo handling work is required at the container terminal. Various causes are decrement of future labor force population by an increasing trend of container moving amount and declining birthrate and aging population. Therefore, this study presents the relative position of hanger and container measurement technology using Single Shot Multibox Detector (SSD) for the purpose of improvement of cargo handling work efficiency and unmanned container terminal. In the case of undetected by SSD, it will be detected using AKAZE feature. The proposed method is applied to 368 images of container gripping taken by a camera installed in a container crane. As a result, Interaction of Union (IoU) targeted for container gripping is 87.79%, and a detection rate is 94.57%.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060451358&origin=inward

Detection of abnormal shadows on temporal subtraction images based on multi-phase CNN 査読有り国際誌

Nagao M., Miyake N., Yoshino Y., Lu H., Kim H., Murakami S., Aoki T., Kido S.

International Conference on Control, Automation and Systems 2018-October 1333 - 1337 2018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（国際会議プロシーディングス）

© ICROS. Recently, visual screening based on CT images become useful tools in the medical fields. However, due to the large number of images and the complexity of the image processing algorithms, image processing technique for the high screening quality is still required. To overcome this problem, some computer aided diagnosis (CAD) algorithms are proposed. Cancer is a leading cause of death both in Japan and worldwide. Detection of cancer region in CT images is the most important task to early detection and early treatment. We have designed and developed a framework combining machine learning based on multi-phase convolutional neural networks (CNN) and temporal subtraction techniques based on non-rigid image registration algorithm. Our main classification method can be built into three main steps; i) preprocessing for image segmentation, ii) image matching for registration, and iii) classification of abnormal regions based on machine learning algorithms. We performed our proposed technique to 25 thoracic MDCT sets and obtained true positive rates of 93.55%, false positive rates of 10.93 /case.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060434756&origin=inward

　前のページ - 次のページ　

<神谷 亨>

論文 - 神谷 亨

<神谷　亨>

論文 - 神谷　亨