Background: Coronavirus disease (COVID-19) has spread explosively worldwide since the beginning of 2020. According to a multinational consensus statement from the Fleischner Society, computed tomography (CT) is a relevant screening tool due to its higher sensitivity for detecting early pneumonic changes. However, physicians are extremely occupied fighting COVID-19 in this era of worldwide crisis. Thus, it is crucial to accelerate the development of an artificial intelligence (AI) diagnostic tool to support physicians. Objective: We aimed to rapidly develop an AI technique to diagnose COVID-19 pneumonia in CT images and differentiate it from non–COVID-19 pneumonia and nonpneumonia diseases. Methods: A simple 2D deep learning framework, named the fast-track COVID-19 classification network (FCONet), was developed to diagnose COVID-19 pneumonia based on a single chest CT image. FCONet was developed by transfer learning using one of four state-of-the-art pretrained deep learning models (VGG16, ResNet-50, Inception-v3, or Xception) as a backbone. For training and testing of FCONet, we collected 3993 chest CT images of patients with COVID-19 pneumonia, other pneumonia, and nonpneumonia diseases from Wonkwang University Hospital, Chonnam National University Hospital, and the Italian Society of Medical and Interventional Radiology public database. These CT images were split into a training set and a testing set at a ratio of 8:2. For the testing data set, the diagnostic performance of the four pretrained FCONet models to diagnose COVID-19 pneumonia was compared. In addition, we tested the FCONet models on an external testing data set extracted from embedded low-quality chest CT images of COVID-19 pneumonia in recently published papers. Results: Among the four pretrained models of FCONet, ResNet-50 showed excellent diagnostic performance (sensitivity 99.58%, specificity 100.00%, and accuracy 99.87%) and outperformed the other three pretrained models in the testing data set. In the additional external testing data set using low-quality CT images, the detection accuracy of the ResNet-50 model was the highest (96.97%), followed by Xception, Inception-v3, and VGG16 (90.71%, 89.38%, and 87.12%, respectively). Conclusions: FCONet, a simple 2D deep learning framework based on a single chest CT image, provides excellent diagnostic performance in detecting COVID-19 pneumonia. Based on our testing data set, the FCONet model based on ResNet-50 appears to be the best model, as it outperformed other FCONet models based on VGG16, Xception, and Inception-v3.

Atrial fibrillation (AF) affects three to five million Americans and is associated with significant morbidity and mortality. Existing methods to diagnose this paroxysmal arrhythmia are cumbersome and/or expensive. We hypothesized that an iPhone 4S can be used to detect AF based on its ability to record a pulsatile photoplethysmogram signal from a fingertip using the built-in camera lens. To investigate the capability of the iPhone 4S for AF detection, we first used two databases, the MIT-BIH AF and normal sinus rhythm (NSR) to derive discriminatory threshold values between two rhythms. Both databases include RR time series originating from 250 Hz sampled ECG recordings. We rescaled the RR time series to 30 Hz so that the RR time series resolution is 1/30 (s) which is equivalent to the resolution from an iPhone 4S. We investigated three statistical methods consisting of the root mean square of successive differences (RMSSD), the Shannon entropy (ShE) and the sample entropy (SampE), which have been proved to be useful tools for AF assessment. Using 64-beat segments from the MIT-BIH databases, we found the beat-to-beat accuracy value of 0.9405, 0.9300, and 0.9614 for RMSSD, ShE, and SampE, respectively. Using an iPhone 4S, we collected 2-min pulsatile time series from 25 prospectively recruited subjects with AF pre- and postelectrical cardioversion. Using derived threshold values of RMSSD, ShE and SampE from the MIT-BIH databases, we found the beat-to-beat accuracy of 0.9844, 0.8494, and 0.9522, respectively. It should be recognized that for clinical applications, the most relevant objective is to detect the presence of AF in the data. Using this criterion, we achieved an accuracy of 100% for both the MIT-BIH AF and iPhone 4S databases.

Sleep disorders are a common affliction for many people even though sleep is one of the most important factors in maintaining good physiological and emotional health. Numerous researchers have proposed various approaches to monitor sleep, such as polysomnography and actigraphy. However, such approaches are costly and often require overnight treatment in clinics. With this in mind, the research presented here has emerged from the question: “Can data be easily collected and analyzed without causing discomfort to patients?” Therefore, the aim of this study is to provide a novel monitoring system for quantifying sleep quality. The data acquisition system is equipped with multimodal sensors, including a three-axis accelerometer and a pressure sensor. To identify sleep quality based on measured data, a novel algorithm, which uses numerous physiological parameters, was proposed. Such parameters include non-REM sleep time, the number of apneic episodes, and sleep durations for dominant poses. To assess the effectiveness of the proposed system, three participants were enrolled in this experimental study for a duration of 20 days. From the experimental results, it can be seen that the proposed monitoring system is effective for quantifying sleep quality.

As mounting evidence suggests a higher incidence of adverse consequences, such as disruption of the immune system, among patients with a history of COVID-19, we aimed to investigate post-COVID-19 conditions on a comprehensive set of allergic diseases including asthma, allergic rhinitis, atopic dermatitis, and food allergy. We used nationwide claims-based cohorts in South Korea (K-CoV-N; n = 836,164; main cohort) and Japan (JMDC; n = 2,541,021; replication cohort A) and the UK Biobank cohort (UKB; n = 325,843; replication cohort B) after 1:5 propensity score matching. Among the 836,164 individuals in the main cohort (mean age, 50.25 years [SD, 13.86]; 372,914 [44.6%] women), 147,824 were infected with SARS-CoV-2 during the follow-up period (2020−2021). The risk of developing allergic diseases, beyond the first 30 days of diagnosis of COVID-19, significantly increased (HR, 1.20; 95% CI, 1.13−1.27), notably in asthma (HR, 2.25; 95% CI, 1.80−2.83) and allergic rhinitis (HR, 1.23; 95% CI, 1.15−1.32). This risk gradually decreased over time, but it persisted throughout the follow-up period (≥6 months). In addition, the risk increased with increasing severity of COVID-19. Notably, COVID-19 vaccination of at least two doses had a protective effect against subsequent allergic diseases (HR, 0.81; 95% CI, 0.68−0.96). Similar findings were reported in the replication cohorts A and B. Although the potential for misclassification of pre-existing allergic conditions as incident diseases remains a limitation, ethnic diversity for evidence of incident allergic diseases in post-COVID-19 condition has been validated by utilizing multinational and independent population-based cohorts.

Wearable reflectance-type photoplethysmography (PPG) sensors, in the form of a band or watch, have recently gained significant attention for real-time heart rate (HR) monitoring applications. To accurately estimate the HR, even during intensive exercise, much effort has focused on simultaneously measured accelerometer signals that cancel out motion artifacts (MAs) from the PPG signal. However, the accelerometer does not always measure true MAs, as it measures not only the acceleration of motion but also gravitational acceleration corresponding to each axis of the accelerometer. In addition, the measured acceleration of the motion itself does not always coincide with the true motion. In this paper, we address the limitations of accelerometers and present an approach using a gyroscope to overcome the issues described. To our knowledge, the limitations of the acceleration signals for MA cancellation have not been addressed in-depth. Here we describe our developed wearable reflectance-type PPG sensor and compare the HR estimation performance between using a gyroscope and an accelerometer from 24 subjects. Our results showed that the gyroscope-assisted approach exhibited better performance than the accelerometer-assisted approach, especially for exercise involving walking.

Some data suggest a higher incidence of diagnosis of autoimmune inflammatory rheumatic diseases (AIRDs) among patients with a history of COVID-19 compared with uninfected patients. However, these studies had methodological shortcomings. To investigate the effect of COVID-19 on long-term risk for incident AIRD over various follow-up periods. Binational, longitudinal, propensity-matched cohort study. Nationwide claims-based databases in South Korea (K-COV-N cohort) and Japan (JMDC cohort). 10 027 506 Korean and 12 218 680 Japanese patients aged 20 years or older, including those with COVID-19 between 1 January 2020 and 31 December 2021, matched to patients with influenza infection and to uninfected control patients. The primary outcome was onset of AIRD (per appropriate codes from the International Classification of Diseases, 10th Revision) 1, 6, and 12 months after COVID-19 or influenza infection or the respective matched index date of uninfected control patients. Between 2020 and 2021, among the 10 027 506 Korean participants (mean age, 48.4 years [SD, 13.4]; 50.1% men), 394 274 (3.9%) and 98 596 (0.98%) had a history of COVID-19 or influenza, respectively. After propensity score matching, beyond the first 30 days after infection, patients with COVID-19 were at increased risk for incident AIRD compared with uninfected patients (adjusted hazard ratio, 1.25 [95% CI, 1.18 to 1.31]) and influenza-infected control patients (adjusted hazard ratio, 1.30 [CI, 1.02 to 1.59]). The risk for incident AIRD was higher with more severe acute COVID-19. Similar patterns were observed in the Japanese cohort. Referral bias due to the pandemic; residual confounding. SARS-CoV-2 infection was associated with increased risk for incident AIRD compared with matched patients without SARS-CoV-2 infection or with influenza infection. The risk for incident AIRD was higher with greater severity of acute COVID-19. National Research Foundation of Korea.