We evaluated the thermal uroflow sensor characteristics of a newly designed cloud networked non-contact uroflowmeter system. The system is roughly divided into a sensor unit and a network unit. The sensor unit consists of four non-contact matrix temperature sensors, a multiplexer, and a built-in microcomputer with WiFi. These electrical devices are built into a raised toilet seat. The network unit is composed of a router, a MVNO line, a database server, and a web server. The acquired data were collected by the database server on the network via WiFi and VPN routers. As a preliminary study, water of 37ºC was tested as pseudo-urine. We obtained the data under various conditions such as different volumes of voided pseudo-urine; different pseudo-urine flow rates; and different positions of falling pseudo-urine in the x-axis, y-axis and angle in the z-axis. The observed data showed a linear relationship with the reference data. The data of pseudo-urine flow rate obtained followed the dynamic change of the reference flow rate. The body weight reduction accompanying 25 times of urination in 5 healthy males showed a linear relationship with the estimated voided volume.

This study addressed the difficulties in using care technologies for executing direct care activities (such as mobility, bathing, meal, and transfer assistance) that require a large amount of time for the care workers. We identified the characteristics of value ranking for difficulties with care technologies by the type of elderly welfare facilities. A total of 1701 care workers of 108 facilities (including nursing homes for the elderly, long-term care beds, long-term care hospitals, multifunctional long-term care in a small group, daycare services for the elderly, and group homes for people with dementia) in X prefecture, who participated in this survey were analyzed. This survey was a self-administered questionnaire ranking the value of care technologies, and the responses were analyzed using Thurston’s pair comparison method. Value ranking by care workers indicated that technologies used to address difficulties in meal assistance, aspiration prevention, and bathing assistance were highly valued, regardless of the type of facility. These results will become the basic data to determine targets to develop nursing care robots, which prioritize the needs of care facilities.

A laser-induced liquid jet (LILJ) was developed as a therapeutic device to achieve a balance between crushing pituitary tumors and preserving fine blood vessels and nerves. The LILJ was previously designed to exscind pituitary tumors on cranial nerves during surgery. Here, we improved the LILJ for another surgical application while maintaining its key advantage of tissue selectivity. We modified the LILJ by incorporating the bubble-jet phenomenon to improve the crushing power of the tool. To allow air into the jet, three types of applicators were manufactured by creating a small hole 0.3mm in diameter in the thin metal tube of the LILJ applicator at a position of 3, 6, or 12mm away from the tip. We also investigated the effects of replacing the thin metal tube with a glass tube having the same small hole. Injecting the liquid jet confirmed the occurrence of the bubble-jet phenomenon. Next, the bubble jet was injected into gelatin, which simulated living tissue, and the crushing effect and depth were measured using a high-speed camera. The results showed that the penetration depth exceeded that of existing instruments when the small hole was 3mm from the tip. Moreover, the crushing power was improved as the small hole was closer from the tip. To further test the applicator with improved crushing power, the jet was injected into an excised pig brain. The bubble jet did not damage the brain parenchyma, and the blood vessels and pia matter on the surface of the brain tissue were preserved. These findings indicate that the tissue-crushing capability is enhanced and the ability to preserve the micro-vessels is retained in the proposed bubble-jet-type LILJ.

We report a segmentation process for multiple anatomical structures in chest X-ray photographs (CXPs) by deep neural networks and the corresponding evaluation results. The segmentation process is a key element of the computer-aided diagnosis (CAD) system, based on changes in appearance of anatomical structures in CXPs. Mainstream, conventional CXP-CAD technologies detect lesions that are machine-learned in advance. However, in a CXP, multiple anatomical structures are depicted in an overlapping manner. Furthermore, if a lesion overlaps with those anatomical structures, it becomes difficult to detect the lesion using conventional methods. Therefore, a new type of CAD system is needed. We use U-Net for the segmentation process. Segmentation targets comprise nine small regions including anatomical structures and boundary lines between anatomical structures. For experimental data assessment, 684 normal cases and 61 abnormal cases were used. For normal cases, Dice coefficients for various structures ranged from 0.653 to 0.919 when using U-Net. For abnormal cases, qualitative evaluation suggested the possibility of anomaly detection. In the future, we will develop anomaly detection for anatomical structures and estimation of abnormal findings for the entire CXP.

This study aims to develop a non-contact uroflowmeter. The sensor unit consists of four non-contact matrix temperature sensors, a multiplexer and a built-in microcomputer. The sensor unit was installed in an existing toilet. Using pure water to pseudo-urine, we examined both change of room temperature and pseudo-urine temperature. The conversion coefficient for estimating pseudo voided volume was affected by both temperatures. Next, we studied how the sitting down on the toilet and flushing influenced the estimated urine volume with subjects. The estimated volume from the temperature sensor showed a good relationship to weight difference before and after the urination.

Tomosynthesis is a technique of limited-angle tomography, which improves the detectability of pulmonary nodules compared to conventional X-ray radiography and acquires images with lower exposure dose than CT. Conventional tomosynthesis requires 10 seconds to obtain 60 projections because of the mechanical motion, and it is difficult for patients to hold the breath. We have proposed 4-projection tomosynthesis that allows acquisition of images in less than 1 second and significant reduction of radiation exposure. In this study, the optimal total scan angle for 4-projection tomosynthesis was examined. A chest phantom with artificial nodules was imaged every 1 degree, and the tomosynthesis images were reconstructed using various projection angles by Back Projection. To assess adequate parameters for image data acquisition, quantitative measures such as contrast noise ratio (CNR) and artifact spread function (ASF) were used to analyze tomosynthetic images. In 4-projection tomosynthesis, narrower projection angle was associated with greater CNR, while depth resolution tended to improve with wider projection angle. Thus, the CNR and depth resolution are in a trade-off relationship. In 4-projection tomosynthesis, the optimum total scan angle was considered to be less than 24 degrees due to the effect of the slice interval used in clinical systems.