小橋 昌司

投球動作の運動学的パラメータから、投手がどの競技レベルに属するかを識別することは、投手の技術的特徴を明らかにすることに繋がる。そこで、本研究では投球動作の計測データを用いて、競技レベルをクラスとみなしたクラス識別と、それに関与する因子の抽出を目的とする。まず、体表に貼付した多数点の赤外線反射マーカの3次元座標時系列データから、投球動作の運動学的指標を算出する。そして、競技レベルを中学、高校、大学、プロの4グループ、算出した指標を入力データとし、機械学習法の一つであるランダムフォレストを用いて競技レベルの推定および推定に関しての各指標の重要度を算出する。本研究では対象を野球投手180名(年齢範囲9〜26歳)とし、144人を学習データ、36人を評価データとした。競技レベルの推定精度は、学習データに対し100%、評価データに対し61.1%であった。また、前腕部の最高速度が競技レベルの識別に大きく関与していることが分かった。(著者抄録)

© 2020, The Author(s). Computer-assisted analysis of dental radiograph in dentistry is getting increasing attention from the researchers in recent years. This is mainly because it can successfully reduce human-made error due to stress, fatigue or lack of experience. Furthermore, it reduces diagnosis time and thus, improves overall efficiency and accuracy of dental care system. An automatic teeth recognition model is proposed here using residual network-based faster R-CNN technique. The detection result obtained from faster R-CNN is further refined by using a candidate optimization technique that evaluates both positional relationship and confidence score of the candidates. It achieves 0.974 and 0.981 mAPs for ResNet-50 and ResNet-101, respectively with faster R-CNN technique. The optimization technique further improves the results i.e. F1 score improves from 0.978 to 0.982 for ResNet-101. These results verify the proposed method’s ability to recognize teeth with high degree of accuracy. To test the feasibility and robustness of the model, a tenfold cross validation (CV) is presented in this paper. The result of tenfold CV effectively verifies the robustness of the model as the average F1 score obtained is more than 0.970. Thus, the proposed model can be used as a useful and reliable tool to assist dental care professionals in dentistry.

© 2020 IEEE. The demand for an automatic bone fracture detection in the emergency section of the hospitals is high for quick diagnosis while maintaining the quality. Previous studies on fracture detection with computed tomography (CT) images or X-ray images have a performance limitation because those methods are based on 2-D image analysis and cannot consider the 3-D internal structure of pelvic bones. This study proposes an automated bone fracture detection from 3-D CT images. Firstly, it introduces a new 3-D annotation method of fractures (called 3-D surface annotation). By using 3-D shape data of pelvic surfaces, it decreases the annotation load significantly. The proposed method estimates the degree of fracture for each point on the pelvic surface. The degree is estimated by 3-D convolutional neural networks (CNN) using 3-D distribution of CT values inside the pelvic surface. The proposed method was validated by using 103 subjects. The accuracy, precision, recall, and specificity for the test data were 69.5%, 61.1%, 56.4%, and 77.7%, respectively.

© 2020 IEEE. The number of rheumatoid arthritis (RA) patients is about 700,000 in Japan. The modified total sharp score (mTSS) calculated from hand X-ray image is a standard diagnosis method of RA progression but can be time-consuming for physicians. Computer-aided diagnostic (CAD) system is expected to improve the diagnostic quality of RA patients. We have previously proposed a CAD system, which can detect finger joint positions using support vector machine and estimate mTSS using support vector regression. This study improves the finger joint detection accuracy by incorporating with statistical shape model, which statistically models individual variety of spatial relationship among finger joints. Experimental results on radiographic images of the hands of 90 RA patients showed that the finger joints were detected with an accuracy of 94.5%, which is higher than 90.6% of the accuracy with the previous method.

© 2020 by the authors. Replacing the humeral head with an artificial one via surgery is one of the options to treat glenohumeral osteoarthritis. Thus, designing the artificial humeral head is an important step to alter clinical outcomes. In order to design the artificial humeral head, the individual variety of the humeral heads should be investigated. The statistical shape model (SSM) has been attracting considerable attention to grasp 3-D shape variety; however, no method to derive the SSM of humeral heads has been studied. This paper proposes a method to construct an SSM of humeral heads based on the anatomical landmarks in shoulder computed tomography (CT) images. The proposed method consists of three steps: humeral head extraction, position and pose alignment, and finally, principle component analysis. The method was applied to 22 male subjects with leave-one-out cross validation. The proposed method obtained an average Dice coefficient of 0.92 to represent the individual shape using the constructed SSM. According to shape analysis of the humeral head, we found that the thickness of the humeral head was associated with individual characteristics of the humeral head. Therefore, it can be said that this study can provide patient-specific design of an artificial humeral head.

© 2020 IEEE. The procedure of orthopedic surgery is quite complicated, and many kinds of equipment have been used. Operating room nurses who deliver surgical instruments to surgeon are supposed to be forced to incur a heavy burden. There are some studies to recognize surgical phase with convolutional neural network (CNN) in minimally invasive laparoscopic surgery only. Previously, we proposed a computer-aided orthopedic surgery (CAOS)-AI navigation system based on CNN. However, the work propose a method to improve accuracy of phase recognition by considering temporal dependency of orthopedic surgery video acquired from surgeon-wearable video camera. The method estimates current surgical phase by combining both temporal dependency and convolutional-long-short term memory network (CNN-LSTM). Experimental results shows a phase recognition accuracy of 59.9% by the proposed method applied in unicomapartmenatal knee arthroplasty (UKA).

© 2020 IEEE. Currently, artificial humeral heads are mainly designed by scaling the average shape of anatomical data, and the humeral head shape is often designed to approximate a sphere or ellipse. It causes a problem that the range of motion (ROM) of the shoulder is limited with the artificial shoulder joint. Improvement of similarity of artificial shoulder joint with actual one may increase the ROM. For the purpose, we previously proposed a method for constructing a statistical shape model (SSM) of the humeral head that represents the inter-individual variation of the humeral head shape using principal component analysis (PCA). In this study, we propose a method to design the artificial humeral head model using Kmeans++ clustering and PCA. First, PCA is applied to the humeral head shape data of the subjects that are aligned and scaled. Next, Kmeans++ clustering is applied to the obtained PC score distribution map, and they are classified into four clusters. The mean shapes were obtained for each cluster and the models were constructed by changing their scale. From the experimental results, it was shown that the artificial humeral head model more similar to the actual humeral head shape was designed by the proposed method.

In order to diagnose osteoarthritis of the shoulder joint, the 3-D shape of the humerus provides the essential information. Also, the segmented region can be utilized for analyzing individual variety of the 3-D shape between normal and anomaly, etc. Since now, there is no study which automatically segments the humerus region from computed tomography (CT) images. U-Net is a fully-convolutional network architecture, and has been applied to some image segmentation problems. This research introduces U-Net architecture to automatically segment the humerus region in shoulder CT images. To validate the proposed method, it has been applied to 19 male subjects. The method achieved 0.946 Dice coefficient, which demonstrates that it successfully segmented the humerus region with a high level of accuracy and precision.

© The Author(s) 2020. Background: During baseball pitching, a high amount of elbow varus torque in the arm cocking-to-acceleration phase is thought to be a biomechanical risk factor for medial elbow pain and injury. The biomechanics of the stride phase may provide preparation for the arm cocking-to-acceleration phase that follows it. Purpose: To determine the kinematic parameters that predict peak elbow varus torque during the stride phase of pitching. Study Design: Descriptive laboratory study. Methods: Participants were 107 high school baseball pitchers (age range, 15-18 years) without shoulder or elbow problems. Whole-body kinematics and kinetics during fastball pitching were analyzed using 3-dimensional measurements from 36 retroreflective markers. A total of 26 kinematic parameters of the upper and lower limbs during the stride phase leading up to the stride foot contact were extracted for multiple regression analysis to assess their combined effect on the magnitude of peak elbow varus torque. Results: Increased wrist extension, elbow pronation, knee flexion on the leading leg, knee extension on the trailing leg at stride foot contact, and upward displacement of the body’s center of mass in the stride phase were significantly correlated with decreased peak elbow varus torque (all P <.05). Moreover, 38% of the variance in peak elbow varus torque was explained by a combination of these 5 significant kinematic variables (P <.001). Conclusion: We found that 5 kinematic parameters during the stride phase and the combination of these parameters were associated with peak elbow varus torque. The stride phase provides biomechanical preparation for pitching and plays a key role in peak elbow varus torque in subsequent pitching phases. Clinical Relevance: The present data can be used to screen pitching mechanics with motion capture assessment to reduce peak elbow varus torque. Decreased peak elbow varus torque is expected to reduce the risk of elbow medial pain and injury.

© 2020, Bentham Science Publishers. All rights reserved. Background: This study presents a novel method of constructing a spatiotemporal statistical shape model (st-SSM) for adult brain. St-SSM is an extension of statistical shape model (SSM) in the temporal domain which will represent the statistical variability of shape as well as the temporal change of statistical variance with respect to time. Aims: Expectation-Maximization (EM) based weighted principal component analysis (WPCA) using a temporal weight function is applied where the eigenvalues of each data are estimated by Estep using temporal eigenvectors, and M-step updates Eigenvectors in order to maximize the variance. Both E and M-step are iterated until updating vectors reaches the convergence point. A weight parameter for each subject is allocated in accordance with the subject’s age to calculate the weighted variance. A Gaussian function is utilized to define the weight function. The center of the function is a time point while the variance is a predefined parameter. Methods: The proposed method constructs adult brain st-SSM by changing the time point between minimum to maximum age range with a small interval. Here, the eigenvectors changes with aging. The feature vector of representing adult brain shape is extracted through a level set algorithm. To validate the method, this study employed 103 adult subjects (age: 22 to 93 y.o. with Mean ± SD = 59.32±16.89) from OASIS database. st-SSM was constructed for time point 40 to 90 with a step of 2. Results: We calculated the temporal deformation change between two-time points and evaluated the corresponding difference to investigate the influence of analysis parameter. An application of the proposed model is also introduced which involves Alzheimer’s disease (AD) identification utilizing support vector machine.

© 2020 SPIE. Surgical tools detection for intraoperative surgical navigation system is essential for better coordination among surgical team in operating room. Because Orthopaedic surgery (OS) differs from laparoscopic, due to a large variety of surgical instruments and techniques making its procedures complicated. Compared to usual object detection in natural images, OS video images are confounded by inhomogeneous illumination; it is hard to directly apply existing studies that are developed for others. Additionally, acquiring Orthopaedic surgery videos is difficult due to recording of surgery videos in restricted surgical environment. Therefore, we propose a deep learning (DL) approach for surgery tools detection in OS videos by integrating knowledge of diverse representative surgery and non-surgery images of tools into the model using transfer learning (TL) and data augmentation. The proposed method has been evaluated for five surgical tools using knee surgery images following 10-fold cross validation. It shows, proposed model (mAP 62.46%) outperforms over conventional model (mAP 60%).

© 2020, Bentham Science Publishers. All rights reserved. Background: Anterior cruciate ligament (ACL) injury causes knee instability which affects sports activity involving cutting and twisting motions. The ACL reconstruction surgery replaces the damaged ACL with artificial one which is fixed to the bone tunnels opened by the surgeon. The outcome of the ACL reconstruction is strongly related to the placement of the bone tunnels, therefore, the optimization of tunnel drilling technique is an important factor to obtain satisfactory surgical results. Aims: The quadrant method is used for the post-operative evaluation of the ACL reconstruction surgery, which evaluates the bone tunnel opening sites on the lateral 2-D X-ray radiograph. Methods: For the purpose of applying the quadrant method to the pre-operative knee MRI, we have synthesized the pseudo lateral 2-D X-ray radiograph from the patients’ knee MRI. This paper proposes a computer-aided surgical planning system for the ACL reconstruction. The proposed system estimates appropriate bone tunnel opening sites on the pseudo lateral 2-D X-ray radiograph synthesized from the pre-operative knee MRI. Results: In the experiment, the proposed method was applied to 98 subjects including subjects with osteoarthritis. The experimental results showed that the proposed method can estimate the bone tunnel opening sites accurately. The other experiment using 36 healthy patients showed that the proposed method is robust to the knee shape deformation caused by disease. Conclusion: It is verified that the proposed method can be applied to subjects with osteoarthritis.

© 2019 IEEE. In recent years, development and spread of medical imaging devices such as X-ray Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) have facilitated acquisition of medical images with low invasiveness and high resolution. At present, some research is being actively conducted to segment a specific area in the image by performing image analysis using the medical image data. In this study, we proposed a method to segment the abdominal organ area using U-Net for cancer radiotherapy support. We investigated segmentation performance of two methods; the first method segments the bladder, the prostate and the rectum, individually, and the second method segments their composite region. As the results, the first method achieved the better performance, and the Dice coefficient was 0.96 on average.

© 2019 IEEE. Uterine peristalsis, which occurs in waves in the uterine region, is one of the fundamental behaviours of uterus in a non-pregnant woman. There are two types of waves in uterine peristalsis; one propagates from the cervix, and the other one does from the fundus. Cine MR images can investigate the wave like uterine peristalsis. Hence, the goal of this study is to quantify the number of peristalsis propagated from the cervix or the fundus using the cine MR images. The proposed method is based on image registration and frequency analysis. The method quantifies the uterine peristalsis by analyzing the frequency spectrum of waves at the cervix and at the fundus individually. The correlation coefficient of the number of peristalsis between visual inspection and the proposed method was 0.9799 at the cervix, and was 0.9999 at the fundus. Thus, this study accurately estimated the number of peristalsis in order to support the diagnosis of the female infertility.

© 2019 IEEE. Human activities like stay, walk, squat, jogging, stair up and down have been estimated by our proposed fuzzy classification system. Data obtained by a wearable biometric sensor device have been used for estimating human activities. The sensor device can obtain electrocardiograms (ECG), 3-axis acceleration, body surface temperature, humidity, ambient temperature, and atmospheric pressure. Calculated body-angle, body vibration, and amount of change of pressure from sensor data have been used for making fuzzy rules. FIR filters have been used for pre-processing of data. Components of the attitude were extracted from the acceleration data. Five-fold and three-fold cross-validation methods have been used when the same person dataset and different person datasets were used for validation, respectively. MHEALTH datasets has also been validated on the fuzzy classification system. Three movements(stay, walk and jogging) have been classified with this system. From the experimental results, in every cases, every movement has been classified with more than 93% accuracy by our proposed method.

© 2019 Wiley Periodicals, Inc. Facial recognition has been employed as a user-friendly person authentication method, and facial spoofing attack has become a common problem. Although there are two kinds of facial spoofing attacks, 2D spoofing and 3D spoofing, almost studies evaluate the performance only for 2D spoofing. Temporal change of face color will be a possible characteristic to detect liveness against to 3D spoofing attack because there is a relationship between the skin blood perfusion change and the temporal color change in facial video images. This paper proposes two features, R-G correlation feature and interarea correlation feature, to detect liveness using video images. Also, liveness detection method using support vector machine is demonstrated. The performance was evaluated by accuracy (ACC) for classifying liveness face and three types of spoofing face—2D printed image, 2D monitor image, and 3D doll. The ACC was 99.2% at the lighting condition of room light, 99.5% at sunlight illuminating the face, and 98.6% at sunlight illuminating the back of the head.

© 2018, Springer-Verlag France SAS, part of Springer Nature. Purpose: To investigate intraoperative kinematics during passive flexion using a surgical navigation system for knees undergoing posterior stabilized (PS) total knee arthroplasty (TKA) with an asymmetric helical post-cam design using navigation system. Methods: In total, 45 knees with both pre- and postoperative kinematic data available were included in the study. Intraoperative kinematic measurements were performed during the course of surgery using the software incorporated in the navigation system. Measurements were performed at the following two time points: (1) before TKA procedure and (2) after TKA implantation. Among the kinematic parameters studied, anterior/posterior translation and axial rotation during flexion were subjected to the analysis. Results: Before surgery, physiologic anterior/posterior translational pattern of the tibia during flexion (rollback of the femur) was found in only 15.6% of the knees. After TKA implantation, postoperative kinematic measurement showed no significant change in the tibial translational during knee flexion. Similarly, with regard to rotation, non-physiologic external tibial rotation in early flexion was observed in the majority of the knees before surgery, and this abnormal kinematic pattern remained after the TKA procedure. Conclusions: The intraoperative three-dimensional motion analysis using a navigation system showed that the physiologic kinematic pattern (anterior translation and internal rotation of the tibia during flexion) of the knee was distorted in osteoarthritic knees undergoing TKA. The abnormal kinematic pattern before surgery was not fully corrected even after implantation of the PS TKA designed to induce natural knee motion; however, no clear relationship between the intraoperative kinematic pattern and knee flexion angle at one year was demonstrated, and the effect of knee kinematics on postoperative knee function and patient’s satisfaction is still unclear.

© 2019 TSI® Press. This work aimed to predict postoperative knee functions of a new patient prior to total knee arthroplasty (TKA) surgery using machine learning, because such prediction is essential for surgical planning and for patients to better understand the TKA outcome. However, the main difficulty is to determine the relationships among individual varieties of preoperative and postoperative knee kinematics. The problem was solved by constructing predictive models from the knee kinematics data of 35 osteoarthritis patients, operated by posterior stabilized implant, based on generalized linear regression (GLR) analysis. Two prediction methods (without and with principal component analysis followed by GLR) along with their sub-classes were proposed, and they were finally evaluated by a leaveone- out cross-validation procedure. The best method can predict the postoperative outcome of a new patient with a Pearson’s correlation coefficient (cc) of 0.84±0.15 (mean±SD) and a root-mean-squared-error (RMSE) of 3.27±1.42 mm for anterior-posterior vs. flexion/extension (A-P pattern), and a cc of 0.89±0.15 and RMSE of 4.25±1.92° for internal-external vs. flexion/extension (i-e pattern). Although these were validated for one type of prosthesis, they could be applicable to other implants, because the definition of knee kinematics, measured by a navigation system, is appropriate for other implants

© 2018 IEEE. In this paper, we propose a new approach to classify ovarian follicles into two classes. A smoothing filter which is designed to consider speckle patterns under the resolution of the ultrasound devices is applied for filtering ovarian follicle images. Then, convolutional neural networks are used for extracting features from the filtered ovarian follicle images. Finally, both features extracted by CNNs from the filtered ovarian follicle images and numerical features defined by our previous works are used for classification. From experimental results, we show the effectiveness of our proposed method.

© 2018 IEEE. Knee implantation is a popular knee surgery to replace damaged knee joint in Total knee arthroplasty (TKA). It is essential to predict postoperative knee kinematic before the surgery for patient-specific TKA surgical planning because outcome of the TKA operation strongly depends on types of prosthesis and surgical methods. Previously, we proposed postoperative kinematics (A-P and i-e patterns) prediction method based on generalized linear regression (GLR). However, this study mainly focuses on comparative performance analysis of the two popular machine learning methods (SVR and NN) in predictive model construction for postoperative kinematics prediction using PCA-based feature extraction, then compared with GLR method. It was found that predictive model's prediction performance slightly varies from each other's because the characteristics features of the kinematic patterns differs from each type. Therefore, this study recommends the best ML method (NN for A-P pattern and GLM for i-e pattern) with high prediction performance for predicting TKA outcome.

© 2018 IEEE. Total knee arthroplasty (TKA) is a surgical procedure to mitigate knee pain and improve functions for people with knee arthritis. The procedure is complicated due to the different surgical tools used in the stages of surgery. Real-time surgical tool recognition can be used to simplify surgical procedures for the surgeon. Also, the presence and movement of tools in surgery are crucial information for the recognition of the operational phase and to identify the surgical workflow. Therefore, this research proposes a real-time system for recognizing surgical tools using a convolutional neural network (CNN). Surgeons wearing smart glasses can see essential information about tools during surgery that may reduce the complication of the procedures. The performance of the proposed method was evaluated by using mean average precision (MAP) with conventional methods which are fast R-CNN and deformable part models. We achieved 87.6% mAP which is better in comparison to the existing methods. With the additional improvements of our proposed method, it can be a future point of reference, also the baseline for operational phase recognition.

© 2018 IEEE. Unusual intense solar flare can have serious impact on the human society. In particular, it may cause serious problems such as damaging electric power plants. It is desirable but difficult to predict intense solar flare because of imbalanced classification problems. To overcome this, we developed Case Based Genetic Algorithm (GA) integrated with Local Optimizer (CBGALO). Here, a Support Vector Machine (SVM) is used as Local Optimizer. However, the prediction precision for learning significantly depends on input data. Therefore, CBGALO was elaborated to extend by a Case Based GA that is able to automatically restart. This forms a good combination searching GA for both learning features and input data (CBRsGcmbGA). Even the currently popular deep learning cannot search the input data for learning automatically or at least evolutionarily. The effect of our approach is proven in predicting X class solar flare as follows: 1) extended CBGALO reached more than 85% of precision in most (12 out of 14) cases and 91.2% at maximum, 2) previous CBGALO reached 84% at most 3) other approaches in the same environment reached less than 75%.

© 2018 IEEE. At present, orthopedic surgery has a large variety of surgical techniques. Procedures are complicated, and many types of equipment have been using in the surgery. So, operating room nurses who deliver surgical instruments to surgeon are supposed to be forced to incur a heavy burden. Although there is a navigation system for assisting surgeons in artificial joint replacement surgery, but no system exists for assisting operating room nurses. This work proposes a computer-aided navigation system that indicates the current procedure and procedure progress for nurses, and also instructs nurses to prepare surgical instruments to be used in the next procedure using smart glasses. Firstly, the system estimates the current status of the surgery procedure using a convolutional neural network (CNN) by utilizing real-time video images taken from smart glasses which was worn by operating surgeon. Then, the system indicates nurses the surgical instrument to be used at the next procedure in the smart glass worn by the nurses. The system was implemented with the object detection technology and the augmented reality. Experiment results demonstrated a satisfactory performance of our proposed system of recognizing surgery procedures.

© 2018 IEEE. The number of Rheumatoid Arthritis (RA) patients increases recently in Japan. Early treatment improves patient&#039;s prognosis and Quality of Life. The appropriate treatment in accordance with RA progression is required for the better prognosis. The hand X-ray image based modified Total Sharp Score (mTSS) is widely used for the diagnosis of RA progression. The mTSS measurement is essential to achieve the appropriate treatment, but its assessment is time consumed. There are some finger joint detection and mTSS estimation methods for the fully automated mTSS measurement, which focus on the mild RA patients. This paper proposes the automatic joint detection method and discusses about the mTSS estimation for the mild-to-severe RA patients. Experimental results on 90 RA patients&#039; hand X-ray images showed that the proposed method detected finger joints with accuracy of 91.8%, and estimated the erosion and JSN score with accuracy of 53.3% and 60.8%, respectively.

© 2018 IEEE. During the early developmental stage, the brain undergoes more changes in size, shape, and appearance than at any other stage in life. A better understanding of brain development can decrease the symptom of development disorder through very early detection and application of remedial education. In this paper, we present a computer-aided diagnosis (CAD) system, which estimates onset probability of brain development disorder using neonatal brain MR images. The CAD system first constructs spatiotemporal statistical shape model (stSSM) of neonatal brain, extracts static and dynamic morphological features, and estimates the probability using machine learning techniques. This paper proposes the stSSM construction method which produces temporally continuous Eigenvectors by extending previous EM-based-stSSM construction method. The method has been validated by applying it to 12 neonatal brains whose revised ages are between - 5 to 730 days.

© 2019 The Institute of Electrical Engineers of Japan. Facial recognition has been employed as a user-friendly person authentication method, and facial spoofing attack has become a common problem. Although there are two kinds of facial spoofing attacks, 2D spoofing and 3D spoofing, almost studies evaluate the performance only for 2D spoofing. Temporal change of face color will be a possible characteristic to detect liveness against to 3D spoofing attack because there is a relationship between the skin blood perfusion change and the temporal color change in facial video images. This paper proposes two features, R-G correlation feature and inter-area correlation feature, to detect liveness using video images. Also, liveness detection method using support vector machine is demonstrated. The performance was evaluated by accuracy (ACC) for classifying liveness face and three types of spoofing face - 2D printed image, 2D monitor image, and 3D doll. The ACC was 99.2% at the lighting condition of room light, 99.5% at sunlight illuminating the face, and 98.6% at sunlight illuminating the back of the head.

本邦には約70万人の慢性関節リウマチ患者が存在し、また毎年数万人が発病する。リウマチは早期治療による予後の著しい改善がみられるが、リウマチの進行度に応じた適切な治療を行う必要がある。リウマチ進行度診断では、年に数回関節レントゲン画像を撮影し、関節破壊進行度mTSスコアを算出しているが、手動であるため膨大な作業時間を要し、また、スコアは主観的評価であるため自動化、定量化の需要が高まっている。本稿では、mTSスコアの自動推定を目的とした手X線画像からの手指関節自動検出法を提案する。また、サポートベクター回帰による手関節X線画像からのmTSスコア推定とその評価を行う。特徴量として関節周辺画素のHOG(histograms of oriented gradient)を用いた。90名のリウマチ患者手X線画像に提案法を適用した結果、81.4%の精度で手指関節を自動認識できた。また、mTSスコア推定結果から、サポートベクター回帰によるmTSスコアの推定が可能であることが示唆された。(著者抄録)

【目的】大腿骨と脛骨両方に原因があり、変形が大きい内反変形膝に対し我々はbi-plane cutの骨切りを大腿骨と脛骨両方に行うdouble level osteotomy(以下、DLO)を行っている。本研究の目的はこのDLO術後の回旋アライメントの変化を3DCTを用いて検討することであった。【方法】26膝を対象とした。回旋アライメントの測定は解析ソフトZiocubeを用い、術前後の大腿骨・脛骨各々の骨座標を一致させて行った。骨切りはともにbi-plane cut、大腿骨はclosing wedge、脛骨はopening wedgeで行い、人工骨はβ-TCPを使用した。統計学的評価はWilcoxonの符号順位検定を用い、危険率が5%未満のものを有意差ありとした。【結果】平均大腿骨遠位回旋角度は術後に内旋2.8°増加と有意な変化を認めた。平均脛骨遠位回旋角度は術後に内旋0.6°増加したが術前後で有意な差は認めなかった。【考察】内反変形膝に対するDLOでは術後、大腿骨で軽度ではあるが遠位骨片の内旋を認めた。この回旋アライメント変化の影響や要因を今後の研究で検討する必要がある。(著者抄録)

【目的】大腿骨と脛骨両方に原因があり、変形が大きい内反変形膝に対し我々はbi-plane cutの骨切りを大腿骨と脛骨両方に行うdouble level osteotomy(以下、DLO)を行っている。本研究の目的はこのDLO術後の回旋アライメントの変化を3DCTを用いて検討することであった。【方法】26膝を対象とした。回旋アライメントの測定は解析ソフトZiocubeを用い、術前後の大腿骨・脛骨各々の骨座標を一致させて行った。骨切りはともにbi-plane cut、大腿骨はclosing wedge、脛骨はopening wedgeで行い、人工骨はβ-TCPを使用した。統計学的評価はWilcoxonの符号順位検定を用い、危険率が5%未満のものを有意差ありとした。【結果】平均大腿骨遠位回旋角度は術後に内旋2.8°増加と有意な変化を認めた。平均脛骨遠位回旋角度は術後に内旋0.6°増加したが術前後で有意な差は認めなかった。【考察】内反変形膝に対するDLOでは術後、大腿骨で軽度ではあるが遠位骨片の内旋を認めた。この回旋アライメント変化の影響や要因を今後の研究で検討する必要がある。(著者抄録)

© 2017 IEEE. Rheumatoid arthritis (RA) damages joints, and the destructed and/or deformed joint causes the pain and reduces the joint function. The prognosis can be improved by early treatment, but it requires accurate evaluation of the degree of RA progression to apply appropriate treatment. The modified total sharp (mTS) score in hand or foot X-ray image has been used to quantitatively evaluate the RA progression evaluation. However, the mTS score measurement takes huge labor and it is very time consuming method because a physician should evaluate progression grade for all hand joints, and the evaluation is subjective. This paper proposes an automated finger joint detection and mTS score estimation method using support vector machine. The experiment in 45 RA patients shows that the proposed method succeeded in detecting the finger joint and estimating the mTS score. As the number of learning data increases, the proposed method can estimate the mTS score with higher accuracy.

© 2017 IEEE. Nurses who engaged in elderly care would like to assess their ability of chewing and swallowing because deterioration of the ability of chewing and swallowing will cause pulmonary aspiration. Currently, nurses can not assess the chewing and swallowing ability quantitatively. In this paper, to quantitatively assess the ability of chewing and swallowing, electromyography (EMG) signals around the lower jaw and the neck are obtained by some electrodes when the subject persons vocalize some Japanese pronunciations. Then, the obtained EMG signals are classified by some machine learning methods. fc-nearest neighbor methods show better classification results for the obtained EMG signals.

Lifestyle and genetics are known to be the major factors causing cerebral aneurysms, but some studies suggest that the shape of cerebral arteries might be correlated with the risk of aneurysm occurrence. This study focuses on the shape of cerebral arteries where cerebral aneurysms tend to occur. First, it extracts the shape feature of the cerebral artery ring, which is a predilection site of cerebral aneurysm, from 3-D magnetic resonance angiography images, and calculates four types of shape feature vectors - 3-D shape, bifurcation angle, degree of meandering, and direction of the branch points. Then, it estimates the risk of cerebral aneurysms occurring, based on the extracted features using support vector machine. To validate the proposed method, we conducted a leave-one-out cross validation test using 80 subjects (40 subjects with and 40 subjects without cerebral aneurysms). The method using a 3-D artery shape achieved 75% sensitivity and 75% specificity; the one using the bifurcation angle showed 47% sensitivity and 41% specificity. The method using the degree of meandering showed 55% sensitivity and 53% specificity, and the one that used the direction of the six branch points showed 30% sensitivity and 27% specificity. These results show that the 3-D artery shape could be a possible indicator for predicting the risk of developing cerebral aneurysms.

This chapter summaries a brain region segmentation method for newborn using magnetic resonance (MR) images. The method deploys fuzzy object growth model (FOGM) which is an extension of fuzzy object model. It is a 4-dimensional model which gives a prior knowledge of brain shape and position at any growing time. First we calculate 4th dimension of FOGM, called growth index in this chapter. Because the growth index will be different from person to person even in the same age group, the method estimates the growth index from cerebral shape using Manifold learning. Using the growth index, FOGM is constructed from the training dataset. To recognize the brain region in evaluating subject, it first estimates the growth index. Then, the brain region is segmented using fuzzy connected image segmentation with the FOGM matched by the growth index. To evaluate the method, this study segments the parenchymal region of 16 subjects (revised age 0–2 years old) using synthesized FOGM.

Implanted knee kinematics recognition is required in total knee arthroplasty (TKA), which replaces damaged knee joint with artificial one. The 3-D kinematics of implanted knee in-vivo is used to quantify the knee function for diagnosis of TKA patients and to evaluate the design of TKA prosthesis and surgical techniques. There are some methods for the implanted knee kinematics estimation, however, those methods are classified into still image analysis. The discontinuous knee kinematics estimated by the still image analysis is not considered as the actual knee kinematics. This paper proposes an kinematics recognition method for implanted knee using particle filter. The proposed method estimates the 3-D pose/position parameters, which are varying in time, based on a priori knowledge of time evolution of the parameters represented by random walk models and utilizing similarity between acquired DR image frame and synthesized DR image based on hypothesized value of the parameters. The experimental results showed that the proposed method successfully estimated the 3-D implanted knee kinematics with an accuracy of 1.61 mm and 0.32°.