Rigid scopes are essential medical devices, yet their lenses incur damage over time due to dirt accumulation and external impacts. This damage, manifesting as cloudy or chipped image screens, can disrupt medical procedures. Traditionally, damage assessment relies on subjective judgment, lacking a quantitative index. To address this, our study developed a quantitative evaluation method for lens damage. A dedicated rigid-scope imaging system was built to ensure a consistent imaging environment. Four images per scope were captured under varying conditions with and without focus adjustment and at different viewing angles and the distance between the objective lens and the object was standardized using white paper. The images were processed in Python with edge enhancement and windowing techniques to emphasize damaged areas. Teacher images highlighting only the damage were created and paired with original images to form a dataset. A supervised machine learning model was developed using 156 training sets, 20 validation sets, and 44 test sets. The model’s performance was evaluated using the Dice coefficient, which ranged from 0.13 to 0.85. Notably, lens cracks and fogging achieved higher Dice scores than dust. These results indicate the model’s utility, and future work will focus on improving dust detection using U-net and larger datasets optimally.

Adequate humidification during mechanical ventilation is essential to protect airway mucosa, maintain mucociliary clearance, and prevent complications such as secretion retention and tube occlusion. Although heated humidifiers and heat and moisture exchangers (HME) are commonly used, only heated humidifiers can reliably deliver gas at physiological humidity levels (37°C, 100% RH, 44 mg/L AH). However, no commercially available system enables real-time, multi-point measurement of temperature and humidity within ventilator circuits. In this study, we developed a novel monitoring system comprising one Breathing Environment Monitoring and Analysis System (BEMAS) and four TH sensors to evaluate temperature and humidity at key points in the ventilator circuit. Four different ventilator models were tested using standardized settings in a controlled environment. Results revealed inter-device variation in delivered temperature and humidity, with absolute humidity levels in the test lung consistently below the physiological target. The system also captured cyclical fluctuations in temperature and humidity corresponding to inspiratory and expiratory phases. These findings highlight the need for better device-specific humidification strategies and suggest the clinical utility of real-time monitoring. Further refinements to sensor size and circuit integration are underway to improve accuracy and applicability in pediatric and adult patients.

In this study, we developed a verification system for cleaning and disinfecting endoscopes by combining an IoT device (equipped with a display) with Hall elements and IR sensors. The effectiveness of each system was evaluated. All systems had the same level of operational accuracy and were able to detect whether the endoscope was connected to the light source device and whether it was hung from a hanger. Moreover, when inappropriate behavior (rehanging a used endoscope on the hanger) occurred, the red LED blinked, the IoT device showed “used” on a blinking red background, and an alarm sounded. When using magnets to mount the endoscope, magnetic sensors were not readily affected by external environments. When it was not possible to attach magnets, the IR sensors were effective. Furthermore, systems using IoT devices presented the cleaning and disinfection status on the display, making it easier for non-medical staff to comprehend the condition of the endoscope. Overall, this contributes to enhancing the safety of endoscope cleaning and disinfection.

This study explores strategic responses that AI-based medical device manufacturers should adopt considering device performance plasticity. We conducted a comparative analysis of AI-based medical devices approved between 2018 and 2023 in Japan and the United States, focusing on the time series changes in the number of approvals, product features and update status, and the attributes of the companies that developed the products. The results showed a greater increase in the number of approvals in the United States than in Japan. In both countries, most of approved products were models without supplemental approval. On the other hand, among products with supplemental approval, Japan tended to have a “function-added” model with multiple different functions, while the United States, with its flexible 510(k) process, showed that many devices adopted an iterative “function-updated” model. Furthermore, foreign companies and startups have actively entered the market. Based on these results, we propose three measures to promote AI-based medical devices development by Japanese medical device manufacturers: (1) strengthening the function-update model, (2) promoting early and continuous dialogue with regulatory authorities, and (3) implementing a short-term product lifecycle that addresses the plasticity of performance, even leading up to approval.

The selection of medical telemeter channels is extremely complex, and the use of incorrect channels can result in severe accidents. Hence, we develop an algorithm for the automatic selection of appropriate medical telemeter channels based on usage regulations. The algorithm is designed to select channels by categorizing addition scenarios based on the operational environment and employing a scoring feature for optimal selection. Furthermore, the algorithm can be integrated into a web application, thus rendering it readily accessible through a browser. Verification tests simulating various operational conditions using a hospital- ward model confirmed the capability of the algorithm for appropriate channel selection with zero incorrect selections. Ultimately, this algorithm is expected to reduce human errors in channel selection and contribute significantly to the safe and efficient channel management of medical telemeters.