Ultrasound is a convenient non-invasive imaging modality used for the diagnosis or detection of various diseases and assessment of therapeutic effects. However, when imaging internal organs, the ultrasound probe must be handled by an operator. The ability to perform hands-free ultrasound imaging of internal organs is likely to offer an unprecedented advantage in various situations such as internal organ monitoring during exercise tests and prolonged monitoring. Toward this end, we have developed a new method of hands-free monitoring using three-dimensional (3D) ultrasound and used this method in portal vein monitoring, which is important for functional evaluation of hepatic and gastrointestinal systems. In previous studies, we developed a hand-made probe holder and used it to capture images of the portal vein, using image tracking and registration to compare the same position of the portal vein. In this study, we first used an abdomen phantom to assess image tracking qualitatively and quantitatively. After validating the method on the phantom, we monitored the portal vein in three healthy subjects using our 3D ultrasound method. Image tracking and registration of the portal veins in three subjects were successfully performed offline. Finally, respiratory analysis and vein diameter measurement were performed based on the image tracking results. The respiratory analysis quantified the respiration-induced portal vein movements. The vein diameter showed changes that might be induced by respiration and heartbeat. These results indicate that our 3D ultrasound method is a potentially useful tool for hands-free monitoring of internal organs.

The ability to fine-tune the movement of swallowing-related organs and change the swallowing pattern to fit the volume of a bolus, texture and the physical properties of the food to be swallowed is referred to as the swallowing reserve. In other words, it is the response capability of food swallowing to avoid choking and aspiration. Herein, we focus on the coordination of the suprahyoid and infrahyoid muscles activities, which are closely related to swallowing movement, as a first step to develop a method to evaluate swallowing reserve, which declines due to neuromuscular disease, muscle weakness caused by aging, to mention a few. First, using two 22-channel electrodes, we measured the surface electromyography (sEMG) signals of suprahyoid and infrahyoid muscles during the following four swallowing conditions: combining two bolus volumes (3 and 15 mL water) and two techniques (normal and effortful swallow). Then, we verified whether the difference in swallowing patterns based on swallowing conditions can be classified from sEMG signals using three machine learning methods; namely, the real-time classification, comprehensive classification, and image recognition method. In the real-time classification method, the mean classification accuracy (MCA) for the four swallowing conditions was as low as 81.5%, indicating that the difference between swallowing conditions performed in a period of approximately 1 s cannot be classified sufficiently by this method. In the comprehensive classification method that applies a majority decision to all the classification results from the start to the end of swallowing, which can be obtained every 16 ms, MCA was 95.1%. Furthermore, in the image recognition method, the change of a series of sEMG signals in the swallowing movement was converted into swallowing pattern image, and the images were classified using a combination of deep convolutional neural networks and support vector machine (SVM). Compared with the comprehensive classification method, the number of training samples for the image recognition method was only 1/26, but the MCA reached 95.7%. This method, which can noninvasively evaluate swallowing patterns that change slightly based on swallowing conditions, could be applied to early detection of reduced swallowing function or a state of frailty (dysphagia potential) in aged individuals.

Assessing blood flow, as well as blood volume and oxygen saturation, is of utmost importance for health monitoring. Assessing microcirculation, as opposed to macrocirculation, is especially important for assessing the local tissue viability. Many optics-based modalities have been developed recently such as laser speckle contrast imaging (LSCI), laser Doppler flowmetry (LDF), diffuse correlation spectroscopy (DCS), and diffuse speckle contrast analysis (DSCA), which all share the advantage of non-invasive measurement of blood flow using non-ionizing radiation. This paper offers a review of these modalities centered around DSCA, which was developed most recently and is considered one of the fastest and most affordable deep tissue perfusion probes.

The micro-strains of a fingernail and toenail caused by pulse wave were measured by a conventional strain measuring method using a biaxial strain gauge. The strain fluctuations of nails were larger in the lateral direction than in the longitudinal direction. In a previous study, pulse wave measurement by a polyvinylidene difluoride (PVDF) film, which is a piezo film, attached to the nail surface was proposed as a new pulse wave measuring method with lower electrical power consumption. The piezo film can transform a nail surface micro-strain caused by pulse wave into an electrical signal. However, the optimal orientation of the piezo film on nails remains unclear. The experimental results of the present study showed that a larger displacement was generated from the nails to the piezo films by aligning the longitudinal direction of the piezo film with the lateral direction of the nails.

The use of social network service (SNS) applications for health communication has revolutionized communication between physicians in recent years. We performed an unrestricted retrospective study focusing on emergency physicians (EPs) in Kyoto University Hospital (KUHP) since timely communication is important during emergencies. EPs used LINE, a popular SNS application in Japan. EPs (n = 22) sent 1752 messages from April 2017 to March 2018. Most messages sent contained text data (82.1%), the remaining contained media (17.9%); media included images (72.6%), LINE stamps (22.9%), LINE albums (2.3%) and files (1.6%). Content analysis by two coders produced 13 categories (n = 1438); these were ‘miscellaneous’, ‘patient’, ‘team’, ‘treatment’, ‘event’, ‘situation’, ‘reference’, ‘announcement’, ‘schedule’, ‘resource’, ‘policy’, ‘transport’ and ‘unknown’. The top five message categories were related to miscellaneous chat (22.5%), patient (19.1%), team (14.3%), treatment (11.8%) and event (6.6%). The largest number of messages among EPs were sent on Monday and Friday. The numbers of messages sent among EPs during day-shift and night-shift were similar. The categories identified influenced our proposal of medical oriented SNS platform features: structured tagging system for messages related to relevant categories (F1); inquiry broadcast system for specific inquiries using structured tagging (F2); image tagging system for images shared within groups (F3) and summarized notifications (F4). Features that need consideration are (1) an opt-in location sharing system between physicians and (2) physicians' access to patient records from the SNS application. In this study, messages discussed by EPs were categorized and the resulting categories influenced our proposal of a physician-centered SNS platform customized to EPs' roles. Since physicians prefer using SNS applications compared to traditional mobile phones, their information needs should be considered. Designing a medical oriented SNS platform that is physician-centered should first include an understanding of topics discussed by physicians. Based on the categories classified, the proposal of physician-centered features for designing a medical oriented SNS platform is also discussed in this paper.

Hyperglycemia is common in critically ill patients and leads to various severe complications and even death. Keeping blood glucose within the range of 80–110 mg/dL (4.4–6.1 mmol/L) has been shown to reduce mortality and morbidity in intensive care units (ICU). Many studies on BG control systems for ICU patients have been reported. However, it is not easy to maintain blood glucose within the desired range because of the time variability of insulin sensitivity in critically ill patients. In this study, to improve the prediction accuracy of blood glucose level in patients, we modified a glycometabolism model developed in our previous study, by identifying parameter values from clinical ICU data. Then, we modified insulin sensitivity online identification algorithm to avoid a sudden change in insulin sensitivity during online identification that updates insulin sensitivity value at intervals of 30 min. Finally, since hypoglycemia prevention as important, we designed a glycemic control system using nonlinear model predictive control based on the modified model and the online identification algorithm of insulin sensitivity. The new glycemic control system achieved 71% of blood glucose measurements within the range of 80–110 mg/dL and 1.5% of measurements below 80 mg/dL, which indicated effectiveness and safety.

EEG source imaging aims to reconstruct the neural activities of the brain accountable for the recorded scalp potentials. This procedure requires solving two problems, namely, forward and inverse problems. For the forward problem, the head is modeled as a volume conductor and the Poisson's equation that describes the relation between neural activities and the observed EEG signals is solved. In this study, we enhanced the forward model by considering the white matter anisotropic conductivity tensor estimated from diffusion-weighted images. The second step is to solve the inverse problem in which the activity of the brain sources is estimated from measured data using the forward solution obtained in the previous step. Spatial filtering, also called beamforming, is an inverse method that reconstructs the time course of the source at a particular location by a linear combination of the sensor space data. We evaluated quantitatively the impact of the enhanced anisotropic forward model on linearly constrained minimum variance beamformer for both superficial and deep sources in a simulation environment, in terms of normalized mean squared error. Results showed that the anisotropic head forward model moderately enhanced the reconstruction of the sources, especially deep thalamic and olfactory sources.

Color information is an important tool for diagnosing melanoma. In this study, we used a hyperspectral imager (HSI), which can measure color information in detail, to develop an automated melanoma diagnosis system. In recent years, the effectiveness of deep learning has become more widely accepted in the field of image recognition. We therefore integrated the deep convolutional neural network with transfer learning into our system. We tried data augmentation to demonstrate how our system improves diagnostic performance. 283 melanoma lesions and 336 non-melanoma lesions were used for the analysis. The data measured by HSI, called the hyperspectral data (HSD), were converted to a single-wavelength image averaged over plus or minus 3 nm. We used GoogLeNet which was pre-trained by ImageNet and then was transferred to analyze the HSD. In the transfer learning, we used not only the original HSD but also artificial augmentation dataset to improve the melanoma classification performance of GoogLeNet. Since GoogLeNet requires three-channel images as input, three wavelengths were selected from those single-wavelength images and assigned to three channels in wavelength order from short to long. The sensitivity and specificity of our system were estimated by 5-fold cross-validation. The results of a combination of 530, 560, and 590 nm (combination A) and 500, 620, and 740 nm (combination B) were compared. We also compared the diagnostic performance with and without the data augmentation. All images were augmented by inverting the image vertically and/or horizontally. Without data augmentation, the respective sensitivity and specificity of our system were 77.4% and 75.6% for combination A and 73.1% and 80.6% for combination B. With data augmentation, these numbers improved to 79.9% and 82.4% for combination A and 76.7% and 82.2% for combination B. From these results, we conclude that the diagnostic performance of our system has been improved by data augmentation. Furthermore, our system succeeds to differentiate melanoma with a sensitivity of almost 80%.

Diabetic retinopathy (DR) is the main cause of retinal damage due to fluid leakage from blood vessels. Automated diagnostic systems assist the ophthalmologists practice manual lesion detection techniques which are tedious and time-consuming. A Diabetic Retinopathy Lesion Discrimination (DRLD) model is proposed for abnormality identification followed by DR lesion detection based on identification of DR pathological symptoms. Shape, intensity and gray-level co-occurrence matrix (GLCM) features are extracted from the identified lesions, and exhaustive statistical analysis is performed for optimal feature selection. Overall accuracies of 97.9% and 91.5% are obtained using multi-layer perceptron neural network (MLPNN) and support vector machine (SVM) classifiers, respectively, for non-diseased versus diseased fundus image discrimination. MLPNN provides better performance for the fundus image discrimination approach, and further accuracy of 98.9% is obtained for DR lesion detection. When compared with other state-of-the-art techniques, the proposed approach provides better performance with significantly less computational complexity. A maximum accuracy improvement of 20.13% in fundus image discrimination and 5.90% in lesion categorization is achieved.

We previously reported the development of a new hybrid medical material comprising bio-based materials with high biocompatibility and artificial materials with characteristics of excellent strength and processability. This material shows sufficient biocompatibility and excellent stability in vivo. Moreover, when applied to the surface of an implantable sensor, the biological reaction on the sensor function surface can be well controlled. For commercialization and widespread use of hybrid materials with such superior properties, sterilization and storage are critical considerations, given that hybrid materials must be processed outside the body prior to application as medical materials in vivo, thus posing a risk of contamination despite best efforts. Therefore, the aim of the present study was to establish an optimal sterilization method that will not impair the biocompatibility of the hybrid material. Toward this end, we tested six sterilization methods for the hybrid material: autoclave (121℃, 20 min), dry heat (160℃, 120 min), ethylene oxide gas (37℃, 120 min), hydrogen peroxide plasma (45℃, 45 min), and gamma ray (25 kGy) with and without lyophilization. After sterilization, the material was cultured with vascular endothelial cells to evaluate the engraftment rate, and was observed with light and scanning electron microscopy to determine shape and structure changes. The results demonstrated that gamma sterilization without lyophilization was the best sterilization method for this material, which preserved the collagen network and showed no change in number of adhered vascular endothelial cells compared to the pre-sterilized material. These findings are useful to promote the commercialization of this hybrid material with combined advantages of synthetic and bio-based materials for widespread clinical application in the engineering of artificial organs.

Neuroscientists usually investigate stimulus selectivity by using a stimulus set and identifying the stimulus that evokes the largest electrophysiological responses averaged over a certain time period. However, the visual environment, and hence the brain activity, changes all the time. A method with sufficiently high temporal resolution for the investigation of dynamic changes in stimulus selectivity is desired. Here, we propose a method by dividing the usual time window for spike rate calculation into multiple smaller time windows. We applied this method to the analysis of temporal change in stimulus selectivity of inferotemporal (IT) cells in macaque monkey recorded previously using microelectrode while they were performing an object discrimination task, in which one object had to be discriminated from others regardless of change in viewing angle. The IT cortex is located at the last stage of the ventral cortical pathway, and is important for object recognition and discrimination. The proposed method theoretically possesses temporal resolution in millisecond order. We demonstrated its ability by following the changes in stimulus selectivity with temporal resolution as high as 20 ms. Furthermore, we divided the response time window into early phase and late phase. In each phase, single cell responses to images (4 objects × 4 views; 16 images in each of the stimulus set) were compared to identify the stimulus evoking the largest response. When comparing the early and late phases, 40% of the cells showed the largest response to the same stimulus (same object and same viewing angle); 13% of the cells showed the largest response to the same object but at different viewing angles; 20% of the cells showed the largest response to different objects at the same viewing angle; and 20% of the cells showed the largest response to different objects at different viewing angles. The dynamic change of stimulus selectivity from early phase to late phase may provide important information about the underlying neuronal mechanism for object recognition. Successful application of the proposed method to the analysis of IT cell activity demonstrates the validity and usefulness of the method.

Breast cancer is one of the most common types of cancer in Japanese women. To address the low spatial resolution challenges associated with mammography and ultrasonography, we focused on the potential of using fluorescence to observe cellular and subcellular structures. Light scattering in living tissue causes a decrease in resolution in in vivo imaging. However, scattering in near-infrared region is weaker than that in the visible region. Therefore, it is essential to investigate the behavior of excitation and emission photons in near-infrared fluorescence within tissues, which could be applied in the detection of breast cancer. We modified our previous multi-layered fluorescence Monte Carlo model of in vivo neuroimaging using quantum dots as the first step for the detection of early-stage breast tumor using both visible and near-infrared light, and developed a model containing skin, breast tissue, and tumor. In the present study, fluorophore concentration and quantum yield parameters were set appropriately based on the mechanism of fluorescence onset. When the depths and sizes of a fluorescent tumor embedded in the breast tissue model were varied, excitation and emission fluence, in addition to intensity were examined from the breast surface. In contrast to visible fluorescence (Ex 488/Em 520), Monte Carlo simulation for breast cancer using near-infrared fluorescence (Ex 780/Em 820) could be used to detect a tumor 1.0 cm in diameter at a depth of 1.0 cm.

The number of people requiring care in Japan has increased with the increase in number of people aged 65 or more. Thus, the production of disposable diapers for the elderly is also increasing. Two types of diapers; pad-type and tape-type, are generally used in combination in nursing homes. Diaper change is a burden for both the caregiver and the care receiver. To reduce the number of diaper changes, this study aimed to develops electrodes as a capacitive sensor attached to the outer surface of the tape-type diaper. The capacitive sensor evaluates the absorption volume of the pad-type diaper, and consists of electrodes and a converter circuit that converts the capacitance of electrodes into an output voltage. The study used a waist-type torso mannequin wearing a pad-type diaper on the inside and a tape-type diaper on the outside. The inside of the torso was filled with saline, and the weight of the torso including the saline was 8.9 kg in total. The electrodes were made of copper tape fixed to an insulating film, and the copper tape was 500-mm long and 25-mm wide with a 20-mm interelectrode gap. The electrodes were attached to the outside of the tape-type diaper and connected to the circuit. Experiment was conducted with the torso placed in supine and 30º lateral positions. Tap water (100 mL) was introduced via a silicone tube between the pad-type diaper and the torso hip surface at a flow rate of 7.8 mL/s, which is the average urinary flow rate for the elderly. This procedure was repeated six times. The experiment was repeated three times. The output voltage increased linearly with increasing volume of water absorbed by the pad-type diaper, in both torso positions. Therefore, the absorption volume was quantitatively evaluated, and the results suggested that the resolution was at least 100 mL when using this sensor.

To develop medical devices, collaborations are necessary between universities and hospitals that have technological problems and companies with the appropriate technological facilities to solve these problems. Large companies do not generally compete in niche markets, where large sales and profits are not expected, which then become the target for small and medium enterprises (SMEs). In this study, the “Mono-Dzukuri region” was considered because of its dense concentration of universities or hospitals and SMEs. However, interventions by industry–academia collaboration coordinators (CDs) are required to ensure smooth collaborations between academia and SMEs that are unfamiliar with medical device development as well as companies that lack market presence in the medical device field. In this study, the Hamamatsu area, a locality that is actively engaged in medical device development through collaborations, was evaluated for different hypotheses. The CDs are integral to forming an appropriate ecosystem for the development team to educate doctors, medical workers, SMEs, and venture companies. Specifically, seminars on medical and clinical research methods are used to educate stakeholders on the importance of collaboration. These seminars provide collaborative opportunities, R&D expenses, and technology transfer and sales promotion support from commercialization to sales. Local governments and banks also support these efforts. CDs from different institutions also actively collaborate with each other; they use co-working spaces and the Coordinator Support System to share information and consult on problems. Through these efforts, CDs have successfully assisted the continuous development of medical devices. Both successful and failed commercialization cases are reported herein. After examining the causes of failure, it was found that collaborations often failed when the needs of the marketing authorization holders could not be met and when the medical and technical needs were not fully understood. To gradually increase the number of successful cases of medical devices developed in Japan, it is necessary to build a development model, expand it to other geographical areas, such as Shinshu, Oita, and Tottori, and support the development of CDs. Additionally, it is important to develop a knowledge base while utilizing the advantages of the local characteristics of these areas.

Chronic diseases such as diabetes, hypertension, and hyperlipidemia increase the medical costs for middle-aged and elderly people, thus requiring preventive intervention. Monitoring, maintaining, or increasing the number of steps walked per day could be expected to effectively reduce medical costs. We evaluated the medical costs for chronic diseases after the implementation of a step count monitoring system with an information communication technology that we hypothesized would efficiently decrease medical costs. We enrolled 342 subjects in the intervention group. An additional 1,025 subjects were selected as controls by matching with the intervention group for age, gender, and total medical costs accumulated in the year prior to the study. The subjects in the intervention group were provided with pedometers that were used with reading devices stationed throughout the city. The duration of the intervention was 18 months. The subjects were given health points based on their activity performance improvements such as the number of steps they walked. The medical costs for the 18 months before and afer the intervention were compared. The effect of cost reduction was evaluated by comparing the percent increase in medical cost from before to after intervention between the control group and the intervention group. After the intervention period, the medical costs increased in the control group but not in the intervention group. The step counts for the first three months and the last three months of the intervention period were assessed, and the results showed that an average step count of 8000 steps was maintained throughout the intervention period. Assuming that the control group showed a natural increase in medical costs, the results suggest that the natural increase in medical costs was suppressed in the intervention group during the intervention period as a result of using the pedometer technology and health points. Therefore, we found that encouraging the participants to proactively participate in walking was an effective strategy.

Ventricular fibrillation (VF) is currently a major cause of sudden cardiac death (SCD). To cure VF, electrical defibrillation is the only therapy. However, strong energy is required. Thus, to reduce the energy or develop a new method is desired. The mechanism of how the electric shock sweeps VF is still controversial. In this article, we summarize evidence and remaining problems of this topic. There are three issues in time sequence of VF: how to initiate, how to continue, and how to terminate. Many investigations to achieve VF-free heart have been reported, but there are currently no definite methods to prevent VF. Thus, to terminate VF is one of the big challenges to prevent SCD. There are two strategies to improve electrical defibrillation: elucidate the substantial mechanism and reduce the energy. (1) Substantial mechanism proposed: In a failed defibrillation episode, at the energy level of the near defibrillation threshold, the initial activation site is related to the repolarization phase of the location. However, it is still not clear whether it is part of the continuous VF activity or initiation of re-VF. It is well known that strong field electric shock (including cathodal and anodal stimuli) has many effects on the cardiac tissue, such as electroporation, virtual electrode effects, and electrophysiological responses, which are influenced by tissue geometry (including fiber orientation and bifurcation of tissues). These phenomena should modify the defibrillation effect. Finally, the characteristics of dynamic spiral wave (SW; the sources of continuity of re-entries) influence the continuity of VF. (2) Efforts to reduce the defibrillation energy: To reduce the defibrillation energy, biphasic pulse, regional cooling, modified stimuli programs, and automated local stimuli to SW are proposed. The superiority of biphasic pulse to monophasic pulse was established in the late 20th century; however, the mechanism is still not well understood. Cooling of some region of the heart ventricles widens the route of SW trajectory and terminates SW. Programming high frequency stimulus or double stimuli according to computer simulation of the heart model could reduce the defibrillation threshold. Automated local stimulus to the site between the tail of SW activation and the next activation front could terminate the SW.

Medial gastrocnemius muscle stiffness was estimated using a system identification technique. The medial gastrocnemius muscle was electrically stimulated using surface Ag-AgCl electrodes and the center of pressure fluctuation in the forward-backward direction was measured with a force plate. Electrically induced fluctuation of the center of pressure was classified according to the displacement and velocity of the center of pressure. The classified steps of displacement and velocity were 0.2 cm and 0.2 cm/s, respectively. The ranges of the classes were ±0.15 cm and ±0.15 cm/s. The classified fluctuations were synchronously averaged and the averaged fluctuation was regarded as an output signal of the transfer function from the electrical stimulation to the fluctuation. The transfer function was identified as an estimate of muscle stiffness using a singular value decomposition method. The average muscle stiffness of eight young male participants ranged from 56.7 to 75.9 N/m. Muscle stiffness was high when the displacement of the center of pressure was positive and the velocity was negative. These characteristics resembled the preceding 0.3 s of the electromyogram. This preceding high muscle activity probably contributed to muscle stiffness. Muscle stiffness was well approximated with a multiple linear regression plane, in which the explanatory variables were the displacement and velocity of the center of pressure.

Subjective, objective, assessment, and plan (SOAP) notes are widely used by physicians to document clinical reasoning in assessing, diagnosing, and treating patients. SOAP notes are also used in medical coding tasks for reimbursement of insurance claims. In Indonesia, medical coders who are independent from physicians assess SOAP notes to assign diagnostic codes and medical procedure codes based on the corresponding International Classification of Diseases standards. Discrepancies between physicians who write the SOAP notes and coders who assign diagnoses and treatments, may occur. These discrepancies were assessed by performing a video-based survey to understand the coder's perspective, allowing the development of a writing support system to achieve unproblematic SOAP notes. This survey found that problematic SOAP notes were not caused by a single problem but by multiple problems. Abbreviations used by physicians are the major problem in assigning diagnostic codes, whereas incomplete data are the major problem in determining planning. This survey also showed that problematic SOAP notes may contain helpful keywords for coders that can help in determining diagnosis and treatment. The findings show that the system should be able to recognize separate sections of the SOAP note to provide writing support features and identify helpful keywords to encourage physicians to write unproblematic SOAP notes.

Being overweight is one of several causes of balance impairment, and it increases the risk of falls. Balance assessments help diagnose this impairment. The outcomes from these assessments are not usually clear to investigate balance impairment in overweight adults. Several methods such as mathematical modeling can be used to investigate the postural control mechanisms in normal balance function. However, there is no study that is focused on the postural control mechanisms in overweight adults. This study aimed to define the postural control models underlying the application of the artificial neural network (ANN) systems in normal weight and overweight populations. Ten participants were recruited and separated into two groups: normal weight (NW) and overweight (OW). There were two processes for determining the postural model in both groups. First, the optimal orders of the nonlinear autoregressive moving average (NARMA) model and the hidden nodes of the ANN system were identified. Mean square error (MSE), Akaike's information criteria (AIC) and residual variance (RV) were used to identify these variables for both groups. Second, the coefficients of these models were defined by the learned weights in the ANN system. The MSE, percent coefficient of variation (%CV), Kolmogorov-Smirnov (KS) test and maximal distance of cumulative distribution function (CDF) were defined to evaluate the performance of the postural models. Furthermore, the orders of the NARMA model and relative importance were utilized to distinguish the postural control mechanisms between the two groups. During the training process, our results indicated that low MSE, AIC and RV were the criteria for hidden nodes and order selection in the NARMA model, which resulted in different patterns of postural models in each group. In the case of the testing process, the findings revealed that the proposed technique could present different postural control strategies for each group. The findings indicated that the postural control mechanism of NW subjects relied on the center of pressure (CoP) in the anterior-posterior (AP) direction, while body sway in the medio-lateral (ML) direction was vital to maintain equilibrium in the OW subjects. Accordingly, the proposed technique could be used to investigate the difference in postural control mechanism between the two groups.

In this study, we attempted to develop a wireless real-time continuous observation device for penile tumescence and/or stiffness for verification of nocturnal penile tumescence (NPT) and for diagnosis and classification of erectile dysfunction (ED). Unlike conventional mechanically wired method, the patient is not constrained by the device and is able to use the device by himself in his own living environment. A soft silicone ring that fits the penis is equipped with a built-in variable inductance, variable frequency oscillator and a one-time use micro-battery. The inductance (hence oscillating frequency) is responsible for the internal stress of the ring, which represents the tumescence and/or stiffness. A non-contact leakage flux coupled to the variable inductor allows proximity telemetry of the phenomenon overnight, or up to the battery life. Our prototype device was successfull in a feasibility test using a desktop model and in a simulated ex vivo test. Now, preparation for pre-clinical trial in healthy volunteer(s) is underway.

Segmentation of computed tomography (CT) images has provided promising methods of constructing precise 3-dimensional heart models. However, the process is labor intensive, because heart regions such as cardiac chambers and blood vessels have similar intensities and exist within a small space. In this paper, we present a tool to efficiently segment cardiac chambers and blood vessels. We extend traditional region growing to be spatially controllable. A user places multiple seeds, each having a bounding area and a threshold, and our tool “grows” regions around each seed independently within its bounding area. To efficiently specify the bounding area, we propose two types of seeds (i.e., sphere and cylinder). We also provide a negative seed that generates fixed background to avoid over-extraction errors. We compared our tool with a traditional scissor tool and confirmed that ours significantly reduced the time required for a segmentation task. We also present segmentation results of CT images of hearts having congenital diseases to illustrate the feasibility of our tool.

Lumbar loading causes increased intervertebral pressure and is an important factor in low back pain. However, it is difficult to quantitatively judge the actions that affect lumbar load and the magnitude of lumbar load increase. Low back pain occurs not only in the workplace but also during activities of daily living. Therefore, it is necessary to investigate the factors inducing low back pain by measuring movements in various planes and determining the magnitude of the lumbar load. Accordingly, the lumbar spine should be examined during various movements. Several studies have examined vertebral bodies in the anteflexion posture. However, the relationship between body flexion angle and vertebral body angle during lateral flexion and rotation remains unknown. In this study, we proposed an estimation method for changes in vertebral body angle during lateral flexion and rotation in the lumbosacral region using a wearable sensor system we previously developed. The accuracy of the proposed estimation method was evaluated and demonstrated using X-ray images.

Information technology has been applied to health management. In our laboratory, we have developed and produced prototypes of new sensors, and attempted to monitor various health parameters noninvasively and unobtrusively. This paper reviews the results of our research related to sensor development. The contents include unobtrusive monitoring in the bed, bath and toilet, and smart house. Then wearable sensors used for photoplethysmography, inertia sensor, deep body temperature measurement, and oxygen uptake monitor are presented. Finally, the requirement of regulatory science is commented

Accurate and trouble-free blood collection reduces the physical and mental risk of patients. Visualization of expert venipuncture techniques is important when conducting blood collection training. Fine motions such as needle insertion for venipuncture should be evaluated by measuring both finger motion and haptic sense. This paper proposes a method for evaluating the needle insertion process in blood collection. Blood collection technique of expert medical staff was measured and analyzed experimentally. A winged blood collection needle was inserted into a forearm model used for venipuncture training. A motion capture system was used to detect finger motions during needle insertion. Furthermore, haptic measurements were made by applying a thin and flexible sensor at the contact interface of the grip to measure both contact pressure and shear stress. The haptic sensor measured contact pressure at the fingers while holding a winged needle, and detected changes in stress components along the directions of needle insertion and skin compression during the venipuncture process. During needle insertion, characteristic changes in stress profiles were observed along with the process of puncturing the component layers of the forearm model. The blood collection procedure could be visualized using haptic sensing throughout the experiment. The proposed haptic sensing system may be useful in enhancing blood collection technique and developing automation of the process.

The most common cancer among women is breast cancer. The survival prognosis and rate of breast cancer differ considerably according to the stage of cancer. Treatment is more efficient if detected early because the progression to a higher stage can be avoided. Digital mammograms are used to identify breast cancer in the early stage. Low contrast mammogram images are usually affected by Poisson noise. In the present work, an improved L₀ gradient minimization smoothing method is proposed. Experiments performed on the DDSM (Digital Database for Screening Mammography) images show that the proposed method is better in sharpening major edges. Further the performance metrics such as peak signal to noise ratio (PSNR), mean square error (MSE) and structural similarity index measure (SSIM) values show that proposed method provides better performance as compared with different existing methods.

In a new Hill-type two-mode model of skeletal muscle, a stretch-evoked force enhancement mode is introduced in addition to a sliding filament mode on the descending limb of the isometric force–length relationship. Dynamic behaviors of muscle on the descending limb are examined in computer simulation using the model. First, force responses of the model to ramp-stretch of muscle are similar to those of frog semitendinosus muscle at different levels of muscle activation. Second, length–velocity phase trajectories of the model in isotonic tetanus contraction are similar to those of rabbit skeletal muscle. Third, firing rates of some motor units of human finger extensor muscle are slightly lower in static position control than in isometric contraction. Computer simulation is performed to reveal mechanisms underlying these behaviors. A hypothesis of static position control on the descending limb based on the simulation results is proposed; i.e., some active motor units of the finger extensor muscle operate in a stretch-evoked force enhancement mode, such that the stable finger position can be maintained owing to the spring-like property of these motor units.

Magnetoencephalography (MEG) has advanced dramatically in the past 50 years, since the first neuromagnetic recording in 1968. Recent MEGs have both high spatial resolution of a few millimeters and high temporal resolution in the order of millisecond. MEG is applied not only clinically, but also in many academic fields including physiology and psychology. In this article, the basic principle and structure of MEG, and the brief history of development are described.

Deep learning is widely used in the development of automatic diagnosis systems for melanoma. However, there are some parameters called hyperparameters which should be set arbitrarily. Optimum setting of hyperparameters is challenging. The dermoscopic images on the database are trained on GoogLeNet. The hyperparameters verified in this study were random seed, solver type, base learning rate, epoch, and batch size. By using a genetic algorithm, these hyperparameters were optimized to obtain higher validation accuracy than other methods such as brute force or Bayesian optimization. The highest validation accuracy was 89.75%. The best hyperparameter settings were: 2 for random seed, RMSProp for solver type, 0.0001 for base learning rate, 30 for epoch, 32 for batch size, and 368 seconds for training time. Using the genetic algorithm, we successfully set the hyperparameters for efficient deep learning. Using the system developed in this study, we plan to search for a broader range of hyperparameters and identify multiple groups including lesions other than melanoma.

Many sensors have been developed to measure physiological information in the toilet for healthcare management and disease prevention. These sensors require personal identification to classify toilet users, assuming that several people use the same toilet. We aimed to provide biometrics based on cardiovascular physiological information obtained from a toilet seat. An electrocardiogram was measured by the toilet seat, and personal identification was performed, as in the previous study. In this study, we developed a ballistocardiographic (BCG) monitoring system using two vibration sensors installed beneath a toilet seat. We recorded 70 BCGs from seven healthy males during their urination. Peaks H, I, J, K, and L of the BCGs were clearly obtained. Twenty-six features such as amplitude and interval were calculated from the BCG peaks. Personal identification was performed using the Mahalanobis distance, and the accuracy of identification and equal error rate (EER) were calculated. The average identification accuracy was found to be 92.2%, and the average EER was found to be 34.1%. Although the EER was not sufficiently low, the average accuracy suggested that the proposed method using a toilet seat could provide adequate biometrics for application in the healthcare system.

Age-related macular degeneration (AMD) causes visual acuity (VA) loss in people aged ≥ 50 years. Common treatments include intravitreal injection of anti-vascular endothelial growth factor agents such as aflibercept. However, lack of response in some patients makes prediction of posttreatment VA difficult. In this paper, we propose a deep neural network model to predict posttreatment VA using pretreatment medical imaging and patient profile data. The proposed model works with image data (optical coherence tomography and color fundus photograph) and patient profile data including gender, age, affected side and pretreatment decimal visual acuity. The model was tested by comparing mean square errors (MSE) between actual and predicted visual acuity obtained from input of image data alone, input of patient profile data alone, and input of both types of data. When examining the concatenation effectiveness of input of both types of data, the outcomes of concatenation conditions 100:100 and 500:500 were compared. For concatenation condition 100:100, MSE was 0.081 for input of image data alone, 0.052 for input of patient profile data alone, and 0.058 for input of both types of data. For concatenation condition 500:500, the MSE values were 0.081, 0.052, and 0.047, respectively. The model proposed provides highly accurate prediction of posttreatment VA and indication of recovery to physicians and patients. The method can handle incomplete images and patient profile data usually collected from patients before treatment.