Effects of Aquatic versus Land High-Intensity Interval Training on Acute Cardiometabolic and Perceptive Responses in Healthy Young Women.

The effects of aquatic high-intensity interval training (AHIIT) on cardiometabolic and perceptive responses when compared to similar land-based exercise (LHIIT) remain unknown. Here, we aimed to (1) establish a matched intensity between mediums and (2) compare the acute cardiometabolic and perceptive responses to the two interventions in healthy young women. Twenty healthy young women performed a stationary running exercise at a matched exercise intensity. The incremental stages, in terms of percentage of heart rate (HR), maximal oxygen uptake (%VO2max), percentage of oxygen uptake reserve (%VO2R), percentage of heart rate reserve (%HRR), and rate of perceived exertion (RPE), were examined and acute cardiometabolic and perceptive responses were evaluated. The results showed that HR was significantly reduced (AHIIT: W 150 ± 19, R 140 ± 18, LHIIT: W 167 ± 16, R 158 ± 16 p < 0.01) and oxygen pulse (AHIIT: W 12 ± 2, R 10 ± 2, LHIIT: W 11 ± 2, R 9 ± 2 p < 0.05) was significantly increased with AHIIT compared to LHIIT. No significant group differences were observed for the perceptive responses. The comparable results demonstrated by the aquatic and land incremental tests allow precise AHIIT and LHIIT prescriptions. AHIIT had distinct differences in HR and oxygen pulse, despite having no distinct difference from LHIIT for some cardiometabolic and affective responses.

Effectiveness of Deep Water Running on Improving Cardiorespiratory Fitness, Physical Function and Quality of Life: A Systematic Review.

Deep Water Running (DWR) is a form of aquatic aerobic exercise simulating the running patterns adopted on dry land. Little is known on the effectiveness of DWR despite gaining popularity. The objective of this study is to systematically review the effects of DWR on cardiorespiratory fitness, physical function, and quality of life in healthy and clinical populations. A systematic search was completed using six databases, including SPORTDiscus, MEDLINE, CINAHL, AMED, Embase, and The Cochrane Library, up to February 2022. Eleven studies evaluating the effectiveness of DWR on cardiorespiratory fitness (CRF), physical function, or quality of life (QoL), compared with no interventions (or standard treatment) or land-based trainings were identified. Data relevant to the review questions were extracted by two independent reviewers when means and standard deviations were reported, and standardized mean differences were calculated. A quality assessment was conducted using selected items from the Downs and Black checklist. A total of 11 clinical trials (7 randomized controlled trials) with a total of 287 participants were included. Meta-analyses were not completed due to the high level of clinical and statistical heterogeneity between studies. Compared with land-based training, DWR showed similar effects on CRF with limited studies reporting outcomes of physical function and QoL compared with the no-exercise control group. DWR appears to be comparable to land-based training for improving CRF. The aquatic environment may provide some advantages for off-loaded exercise at high intensity in populations that are weak, injured or in pain, but more studies are required.

Enablers and barriers of harnessing food waste to address food insecurity: a scoping review.

Despite producing sufficient food for the global population, the growing prevalence of food insecurity in developed countries is cause for concern. The millions of metric tons of food wasted each year could be used instead to drastically lower rates of food insecurity and address food sustainability. In this scoping review, we aimed to identify barriers to and enablers of harnessing food waste across food sectors, including food retail, households, and food rescue organizations, to address food insecurity in a developed country, Australia. The findings demonstrate that research on and responsibility for harnessing food waste for food insecurity has predominantly fallen on ill-equipped food rescue organizations. Three primary policy advancements paramount to harnessing food waste to address food insecurity include (1) improving partnerships and subsidies to minimize transportation costs for redistributing imperfect or surplus food from farmers and retailers to those who with food insecurity; (2) enhancing existing partnerships and subsidies to stably involve more nutrition experts in food rescue organizations to improve the quality of foods being redistributed to those facing food insecurity; and (3) initiating interventions and campaigns that combine the following 5 characteristics: free to the participants; address food literacy; use multiple mass-media tools; are age tailored; and frame messages within personal values.

Unsupervised seizure identification on EEG.

BACKGROUND AND OBJECTIVE: Epilepsy is one of the most common neurological disorders, whose development is typically detected via early seizures. Electroencephalogram (EEG) is prevalently employed for seizure identification due to its routine and low expense collection. The stochastic nature of EEG makes manual seizure inspections laborsome, motivating automated seizure identification. The relevant literature focuses mostly on supervised machine learning. Despite their success, supervised methods require expert labels indicating seizure segments, which are difficult to obtain on clinically-acquired EEG. Thus, we aim to devise an unsupervised method for seizure identification on EEG. METHODS: We propose the first fully-unsupervised deep learning method for seizure identification on raw EEG, using a variational autoencoder (VAE). In doing so, we train the VAE on recordings without seizures. As training captures non-seizure activity, we identify seizures with respect to the reconstruction errors at inference time. Moreover, we extend the traditional VAE training loss to suppress EEG artifacts. Our method does not require ground-truth expert labels indicating seizure segments or manual feature extraction. RESULTS: We implement our method using the PyTorch library and execute experiments on an NVIDIA V100 GPU. We evaluate our method on three benchmark EEG datasets: (i) intracranial recordings from the University of Pennsylvania and the Mayo Clinic, (ii) scalp recordings from the Temple University Hospital of Philadelphia, and (iii) scalp recordings from the Massachusetts Institute of Technology and the Boston Children's Hospital. To assess performance, we report accuracy, precision, recall, Area under the Receiver Operating Characteristics Curve (AUC), and p-value under the Welch t-test for distinguishing seizure vs. non-seizure EEG windows. Our approach can successfully distinguish seizures from non-seizure activity, with up to 0.83 AUC on intracranial recordings. Moreover, our algorithm has the potential for real-time inference, by processing at least 10 s of EEG in a second. CONCLUSION: We take the first successful steps in deep learning-based unsupervised seizure identification on raw EEG. Our approach has the potential of alleviating the burden on clinical experts regarding laborsome EEG inspections for seizures. Furthermore, aiding the identification of early seizures via our method could facilitate successful detection of epilepsy development and initiate antiepileptogenic therapies.

Investigating sleep spindle density and schizophrenia: A meta-analysis.

Sleep abnormalities are an early feature of schizophrenia (SZ) characterized by reductions in sleep spindles that are associated with deficits in brain connectivity and cognitive function. This study investigated sleep spindle density (SSD) differences between SZ, first episode psychosis (FEP), and family high-risk (FHR) populations and matched healthy controls (HC) by investigating recent studies via a meta-analysis. We collected experimental, demographic, and methodological metrics from eligible studies across multiple online databases. 14 total studies survived the inclusion and exclusion criteria for a total of 337 patients and relatives and 339 HC. R-Studio was used to run the meta-analysis via the meta and metaphor packages. A heterogeneity score of I2 = 80% was calculated and thus a random effects model was chosen. We report a large effect size for SSD in patients compared to controls. Furthermore, illness duration was significantly associated with SSD. Our next step to understanding sleep spindles would be to investigate SSD's use as a predictor for SZ or attempt to normalize SSD deficits as a therapeutic option.

COVID-19 Vaccination Dynamics in the US: Coverage Velocity and Carrying Capacity Based on Socio-demographic Vulnerability Indices in California.

Coronavirus disease 2019 (COVID-19) disparities among vulnerable populations are of paramount concern that extend to vaccine administration. With recent uptick in infection rates, dominance of the delta variant, and authorization of a third booster shot, understanding the population-level vaccine coverage dynamics and underlying sociodemographic factors is critical for achieving equity in public health outcomes. This study aimed to characterize the scope of vaccine inequity in California counties through modeling the trends of vaccination using the Social Vulnerability Index (SVI). Overall SVI, its four themes, and 9228 data points of daily vaccination numbers from December 15, 2020, to May 23, 2021, across all 58 California counties were used to model the growth velocity and anticipated maximum proportion of population vaccinated, defined as having received at least one dose of vaccine. Based on the overall SVI, the vaccination coverage velocity was lower in counties in the high vulnerability category (v = 0.0346, 95% CI 0.0334, 0.0358) compared to moderate (v = 0.0396, 95% CI 0.0385, 0.0408) and low (v = 0.0414, 95% CI 0.0403, 0.0425) vulnerability categories. SVI Theme 3 (minority status and language) yielded the largest disparity in coverage velocity between low and high-vulnerable counties (v = 0.0423 versus v = 0.035, P < 0.001). Based on the current trajectory, while counties in low-vulnerability category of overall SVI are estimated to achieve a higher proportion of vaccinated individuals, our models yielded a higher asymptotic maximum for highly vulnerable counties of Theme 3 (K = 0.544, 95% CI 0.527, 0.561) compared to low-vulnerability counterparts (K = 0.441, 95% CI 0.432, 0.450). The largest disparity in asymptotic proportion vaccinated between the low and high-vulnerability categories was observed in Theme 2 describing the household composition and disability (K = 0.602, 95% CI 0.592, 0.612; versus K = 0.425, 95% CI 0.413, 0.436). Overall, the large initial disparities in vaccination rates by SVI status attenuated over time, particularly based on Theme 3 status which yielded a large decrease in cumulative vaccination rate ratio of low to high-vulnerability categories from 1.42 to 0.95 (P = 0.002). This study provides insight into the problem of COVID-19 vaccine disparity across California which can help promote equity during the current pandemic and guide the allocation of future vaccines such as COVID-19 booster shots.

A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play.

The game of chess is the longest-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. By contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go by reinforcement learning from self-play. In this paper, we generalize this approach into a single AlphaZero algorithm that can achieve superhuman performance in many challenging games. Starting from random play and given no domain knowledge except the game rules, AlphaZero convincingly defeated a world champion program in the games of chess and shogi (Japanese chess), as well as Go.

Mastering the game of Go without human knowledge.

A long-standing goal of artificial intelligence is an algorithm that learns, tabula rasa, superhuman proficiency in challenging domains. Recently, AlphaGo became the first program to defeat a world champion in the game of Go. The tree search in AlphaGo evaluated positions and selected moves using deep neural networks. These neural networks were trained by supervised learning from human expert moves, and by reinforcement learning from self-play. Here we introduce an algorithm based solely on reinforcement learning, without human data, guidance or domain knowledge beyond game rules. AlphaGo becomes its own teacher: a neural network is trained to predict AlphaGo's own move selections and also the winner of AlphaGo's games. This neural network improves the strength of the tree search, resulting in higher quality move selection and stronger self-play in the next iteration. Starting tabula rasa, our new program AlphaGo Zero achieved superhuman performance, winning 100-0 against the previously published, champion-defeating AlphaGo.