Machine learning algorithm for early-stage prediction of severe morbidity in COVID-19 pneumonia patients based on bio-signals.

BACKGROUND: Paralysis of medical systems has emerged as a major problem not only in Korea but also globally because of the COVID-19 pandemic. Therefore, early identification and treatment of COVID-19 are crucial. This study aims to develop a machine-learning algorithm based on bio-signals that predicts the infection three days in advance before it progresses from mild to severe, which may necessitate high-flow oxygen therapy or mechanical ventilation. METHODS: The study included 2758 hospitalized patients with mild severity COVID-19 between July 2020 and October 2021. Bio-signals, clinical information, and laboratory findings were retrospectively collected from the electronic medical records of patients. Machine learning methods included random forest, random forest ranger, gradient boosting machine, and support vector machine (SVM). RESULTS: SVM showed the best performance in terms of accuracy, kappa, sensitivity, detection rate, balanced accuracy, and run-time; the area under the receiver operating characteristic curve was also quite high at 0.96. Body temperature and SpO2 three and four days before discharge or exacerbation were ranked high among SVM features. CONCLUSIONS: The proposed algorithm can predict the exacerbation of severity three days in advance in patients with mild COVID-19. This prediction can help effectively manage the reallocation of appropriate medical resources in clinical settings. Therefore, this algorithm can facilitate adequate oxygen therapy and mechanical ventilator preparation, thereby improving patient prognosis, increasing the efficiency of medical systems, and mitigating the damage caused by a global pandemic.

Enhanced DBR mirror design via D3QN: A reinforcement learning approach.

Modern optical systems are important components of contemporary electronics and communication technologies, and the design of new systems has led to many innovative breakthroughs. This paper introduces a novel application based on deep reinforcement learning, D3QN, which is a combination of the Dueling Architecture and Double Q-Network methods, to design distributed Bragg reflectors (DBRs). Traditional design methods are based on time-consuming iterative simulations, whereas D3QN is designed to optimize the multilayer structure of DBRs. This approach enabled the reflectance performance and compactness of the DBRs to be improved. The reflectance of the DBRs designed using D3QN is 20.5% higher compared to designs derived from the transfer matrix method (TMM), and these DBRs are 61.2% smaller in terms of their size. These advancements suggest that deep reinforcement learning, specifically the D3QN methodology, is a promising new method for optical design and is more efficient than traditional techniques. Future research possibilities include expansion to 2D and 3D design structures, where increased design complexities could likely be addressed using D3QN or similar innovative solutions.

Direct visualization of carbon black aggregates in nitrile butadiene rubber by THz near-field microscope.

The use of filling agents for rubber reinforcement is beneficial in various industrial applications, and several experimental methods have been used to study the effect of fillers on rubber. However, due to the lack of a suitable imaging technique, filler dispersion and distribution in rubber cannot be easily displayed. Thus, we utilize the THz near-field microscope (THz-NFM) to directly visualize the distribution of carbon black (CB) aggregates in nitrile butadiene rubber (NBR). The THz time-domain spectroscopy (THz-TDS) was used to evaluate the optical properties of the NBR specimens. Results revealed significant indices contrast between CB and NBR at the THz regime, which was attributed to the variation in electrical conductivities. The micrographs of NBR in the THz-NFM revealed the distribution of CB aggregates. The area fraction (AF) of the CB aggregates was calculated using a binary thresholding algorithm to compare with the transmission electron microscope method. Both methods yielded comparable AF values, suggesting, for the first time, that CB can be detected in the NBR without preprocessing the specimens.

Long-term reliable physical health monitoring by sweat pore-inspired perforated electronic skins.

Electronic skins (e-skins)-electronic sensors mechanically compliant to human skin-have long been developed as an ideal electronic platform for noninvasive human health monitoring. For reliable physical health monitoring, the interface between the e-skin and human skin must be conformal and intact consistently. However, conventional e-skins cannot perfectly permeate sweat in normal day-to-day activities, resulting in degradation of the intimate interface over time and impeding stable physical sensing. Here, we present a sweat pore-inspired perforated e-skin that can effectively suppress sweat accumulation and allow inorganic sensors to obtain physical health information without malfunctioning. The auxetic dumbbell through-hole patterns in perforated e-skins lead to synergistic effects on physical properties including mechanical reliability, conformability, areal mass density, and adhesion to the skin. The perforated e-skin allows one to laminate onto the skin with consistent homeostasis, enabling multiple inorganic sensors on the skin to reliably monitor the wearer's health over a period of weeks.