Dance Exergames for Older Adults: A Systematic Review and Meta-Analysis With Narrative Synthesis.

BACKGROUND: A systematic review and meta-analysis with narrative synthesis was conducted to evaluate the impact of dance exergaming on older adults' health-related outcomes and its feasibility, usability, and safety. METHODS: PubMed, Scopus, CINAHL, Web of Science, The Cochrane Library, ProQuest Dissertations and Theses Global, and Google Scholar were searched from inception to December 7, 2023. Interventional studies using immersive or nonimmersive virtual reality platforms conducted on older adults ≥60 years old were eligible. Meta-analysis was conducted using the random effects model by pooling mean differences (MD) or standardized mean differences. Outcomes were narratively synthesized when meta-analysis was not possible. RESULTS: Forty-three articles from 37 studies were included (n = 1 139 participants at baseline). Postintervention, dynamic balance measured using Berg Balance Scale (pooled MD = 2.65, 95% CI: 1.73-3.57, p < .0001), Timed-Up-and-Go times (pooled MD = -1.04, 95% CI: -2.06 to -0.03, p = .04), choice stepping reaction time (pooled MD = -92.48, 95% CI: -167.30 to -17.67, p = .02), and movement time (pooled MD = -50.33, 95% CI: -83.34 to -17.33, p = .003) were significantly better in the experimental group compared to the control group. Adherence ranged from 76.5% to 100%, whereas attrition ranged from 9.1% to 31.9%. Most participants completed the intervention with no or minimal adverse effects. CONCLUSIONS: Dance exergames are effective, feasible, usable, and safe for older adults. Further research is needed as the findings were limited by small sample sizes. Many studies could not be included in the meta-analysis as outcomes were too varied.

A novel bidirectional LSTM deep learning approach for COVID-19 forecasting.

COVID-19 has resulted in significant morbidity and mortality globally. We develop a model that uses data from thirty days before a fixed time point to forecast the daily number of new COVID-19 cases fourteen days later in the early stages of the pandemic. Various time-dependent factors including the number of daily confirmed cases, reproduction number, policy measures, mobility and flight numbers were collected. A deep-learning model using Bidirectional Long-Short Term Memory (Bi-LSTM) architecture was trained on data from 22nd Jan 2020 to 8 Jan 2021 to forecast the new daily number of COVID-19 cases 14 days in advance across 190 countries, from 9 to 31 Jan 2021. A second model with fewer variables but similar architecture was developed. Results were summarised by mean absolute error (MAE), root mean squared error (RMSE), mean absolute percentage error (MAPE), and total absolute percentage error and compared against results from a classical ARIMA model. Median MAE was 157 daily cases (IQR: 26-666) under the first model, and 150 (IQR: 26-716) under the second. Countries with more accurate forecasts had more daily cases and experienced more waves of COVID-19 infections. Among countries with over 10,000 cases over the prediction period, median total absolute percentage error was 33% (IQR: 18-59%) and 34% (IQR: 16-66%) for the first and second models respectively. Both models had comparable median total absolute percentage errors but lower maximum total absolute percentage errors as compared to the classical ARIMA model. A deep-learning approach using Bi-LSTM architecture and open-source data was validated on 190 countries to forecast the daily number of cases in the early stages of the COVID-19 outbreak. Fewer variables could potentially be used without impacting prediction accuracy.

Inferring cancer disease response from radiology reports using large language models with data augmentation and prompting.

OBJECTIVE: To assess large language models on their ability to accurately infer cancer disease response from free-text radiology reports. MATERIALS AND METHODS: We assembled 10 602 computed tomography reports from cancer patients seen at a single institution. All reports were classified into: no evidence of disease, partial response, stable disease, or progressive disease. We applied transformer models, a bidirectional long short-term memory model, a convolutional neural network model, and conventional machine learning methods to this task. Data augmentation using sentence permutation with consistency loss as well as prompt-based fine-tuning were used on the best-performing models. Models were validated on a hold-out test set and an external validation set based on Response Evaluation Criteria in Solid Tumors (RECIST) classifications. RESULTS: The best-performing model was the GatorTron transformer which achieved an accuracy of 0.8916 on the test set and 0.8919 on the RECIST validation set. Data augmentation further improved the accuracy to 0.8976. Prompt-based fine-tuning did not further improve accuracy but was able to reduce the number of training reports to 500 while still achieving good performance. DISCUSSION: These models could be used by researchers to derive progression-free survival in large datasets. It may also serve as a decision support tool by providing clinicians an automated second opinion of disease response. CONCLUSIONS: Large clinical language models demonstrate potential to infer cancer disease response from radiology reports at scale. Data augmentation techniques are useful to further improve performance. Prompt-based fine-tuning can significantly reduce the size of the training dataset.

Prediction of anticancer peptides based on an ensemble model of deep learning and machine learning using ordinal positional encoding.

Anticancer peptides (ACPs) are the types of peptides that have been demonstrated to have anticancer activities. Using ACPs to prevent cancer could be a viable alternative to conventional cancer treatments because they are safer and display higher selectivity. Due to ACP identification being highly lab-limited, expensive and lengthy, a computational method is proposed to predict ACPs from sequence information in this study. The process includes the input of the peptide sequences, feature extraction in terms of ordinal encoding with positional information and handcrafted features, and finally feature selection. The whole model comprises of two modules, including deep learning and machine learning algorithms. The deep learning module contained two channels: bidirectional long short-term memory (BiLSTM) and convolutional neural network (CNN). Light Gradient Boosting Machine (LightGBM) was used in the machine learning module. Finally, this study voted the three models' classification results for the three paths resulting in the model ensemble layer. This study provides insights into ACP prediction utilizing a novel method and presented a promising performance. It used a benchmark dataset for further exploration and improvement compared with previous studies. Our final model has an accuracy of 0.7895, sensitivity of 0.8153 and specificity of 0.7676, and it was increased by at least 2% compared with the state-of-the-art studies in all metrics. Hence, this paper presents a novel method that can potentially predict ACPs more effectively and efficiently. The work and source codes are made available to the community of researchers and developers at https://github.com/khanhlee/acp-ope/.

Improved Prediction Model of Protein and Peptide Toxicity by Integrating Channel Attention into a Convolutional Neural Network and Gated Recurrent Units.

In recent times, the importance of peptides in the biomedical domain has received increasing concern in terms of their effect on multiple disease treatments. However, before successful large-scale implementation in the industry, accurate identification of peptide toxicity is a vital prerequisite. The existing computational methods have reached good results from toxicity prediction, and we present an improved model based on different deep learning architectures. The modification mainly focuses on two aspects: sequence encoding and variational information bottlenecks. Consequently, one of our modified plans shows an obvious increase in sensitivity, while the rest show good performance meanwhile adding novelty in the peptide toxicity prediction domain. In detail, our best model could achieve an accuracy of 97.38 and 95.03% in protein and peptide toxicity predictions, respectively. The performance was superior to previous predictors on the same datasets.

Predicting Antimalarial Activity in Natural Products Using Pretrained Bidirectional Encoder Representations from Transformers.

Malaria is a threatening disease that has claimed many lives and has a high prevalence rate annually. Through the past decade, there have been many studies to uncover effective antimalarial compounds to combat this disease. Alongside chemically synthesized chemicals, a number of natural compounds have also been proven to be as effective in their antimalarial properties. Besides experimental approaches to investigate antimalarial activities in natural products, computational methods have been developed with satisfactory outcomes obtained. In this study, we propose a novel molecular encoding scheme based on Bidirectional Encoder Representations from Transformers and used our pretrained encoding model called NPBERT with four machine learning algorithms, including k-Nearest Neighbors (k-NN), Support Vector Machines (SVM), eXtreme Gradient Boosting (XGB), and Random Forest (RF), to develop various prediction models to identify antimalarial natural products. The results show that SVM models are the best-performing classifiers, followed by the XGB, k-NN, and RF models. Additionally, comparative analysis between our proposed molecular encoding scheme and existing state-of-the-art methods indicates that NPBERT is more effective compared to the others. Moreover, the deployment of transformers in constructing molecular encoders is not limited to this study but can be utilized for other biomedical applications.

Investigating Psychological Differences Between Nurses and Other Health Care Workers From the Asia-Pacific Region During the Early Phase of COVID-19: Machine Learning Approach.

BACKGROUND: As the COVID-19 pandemic evolves, challenges in frontline work continue to impose a significant psychological impact on nurses. However, there is a lack of data on how nurses fared compared to other health care workers in the Asia-Pacific region. OBJECTIVE: This study aims to investigate (1) the psychological outcome characteristics of nurses in different Asia-Pacific countries and (2) psychological differences between nurses, doctors, and nonmedical health care workers. METHODS: Exploratory data analysis and visualization were conducted on the data collected through surveys. A machine learning modeling approach was adopted to further discern the key psychological characteristics differentiating nurses from other health care workers. Decision tree-based machine learning models (Light Gradient Boosting Machine, GradientBoost, and RandomForest) were built to predict whether a set of psychological distress characteristics (ie, depression, anxiety, stress, intrusion, avoidance, and hyperarousal) belong to a nurse. Shapley Additive Explanation (SHAP) values were extracted to identify the prominent characteristics of each of these models. The common prominent characteristic among these models is akin to the most distinctive psychological characteristic that differentiates nurses from other health care workers. RESULTS: Nurses had relatively higher percentages of having normal or unchanged psychological distress symptoms relative to other health care workers (n=233-260 [86.0%-95.9%] vs n=187-199 [74.8%-91.7%]). Among those without psychological symptoms, nurses constituted a higher proportion than doctors and nonmedical health care workers (n=194 [40.2%], n=142 [29.5%], and n=146 [30.3%], respectively). Nurses in Vietnam showed the highest level of depression, stress, intrusion, avoidance, and hyperarousal symptoms compared to those in Singapore, Malaysia, and Indonesia. Nurses in Singapore had the highest level of anxiety. In addition, nurses had the lowest level of stress, which is the most distinctive psychological outcome characteristic derived from machine learning models, compared to other health care workers. Data for India were excluded from the analysis due to the differing psychological response pattern observed in nurses in India. A large number of female nurses emigrating from South India could not have psychologically coped well without the support from family members while living alone in other states. CONCLUSIONS: Nurses were least psychologically affected compared to doctors and other health care workers. Different contexts, cultures, and points in the pandemic curve may have contributed to differing patterns of psychological outcomes amongst nurses in various Asia-Pacific countries. It is important that all health care workers practice self-care and render peer support to bolster psychological resilience for effective coping. In addition, this study also demonstrated the potential use of decision tree-based machine learning models and SHAP value plots in identifying contributing factors of sophisticated problems in the health care industry.

Improved Prediction Model of Protein Lysine Crotonylation Sites Using Bidirectional Recurrent Neural Networks.

Histone lysine crotonylation (Kcr) is a post-translational modification of histone proteins that is involved in the regulation of gene transcription, acute and chronic kidney injury, spermatogenesis, depression, cancer, and so forth. The identification of Kcr sites in proteins is important for characterizing and regulating primary biological mechanisms. The use of computational approaches such as machine learning and deep learning algorithms have emerged in recent years as the traditional wet-lab experiments are time-consuming and costly. We propose as part of this study a deep learning model based on a recurrent neural network (RNN) termed as Sohoko-Kcr for the prediction of Kcr sites. Through the embedded encoding of the peptide sequences, we investigate the efficiency of RNN-based models such as long short-term memory (LSTM), bidirectional LSTM (BiLSTM), and bidirectional gated recurrent unit (BiGRU) networks using cross-validation and independent tests. We also established the comparison between Sohoko-Kcr and other published tools to verify the efficiency of our model based on 3-fold, 5-fold, and 10-fold cross-validations using independent set tests. The results then show that the BiGRU model has consistently displayed outstanding performance and computational efficiency. Based on the proposed model, a webserver called Sohoko-Kcr was deployed for free use and is accessible at https://sohoko-research-9uu23.ondigitalocean.app.

Development and validation of the SARICA score to predict survival after return of spontaneous circulation in out of hospital cardiac arrest using an interpretable machine learning framework.

BACKGROUND: Accurate and timely prognostication of patients with out-of-hospital cardiac arrest (OHCA) who achieved the return of spontaneous circulation (ROSC) is crucial in clinical decision-making, resource allocation, and communications with next-of-kins. We aimed to develop the Survival After ROSC in Cardiac Arrest (SARICA), a practical clinical decision tool to predict survival in OHCA patients who attained ROSC. METHODS: We utilized real-world Singapore data from the population-based Pan-Asian Resuscitation Outcomes Study between 2010-2018. We excluded patients without ROSC. The dataset was segmented into training (60%), validation (20%) and testing (20%) cohorts. The primary endpoint was survival (to 30-days or hospital discharge). AutoScore, an interpretable machine-learning based clinical score generation algorithm, was used to develop SARICA. Candidate factors were chosen based on objective demographic and clinical factors commonly available at the time of admission. Performance of SARICA was evaluated based on receiver-operating curve (ROC) analyses. RESULTS: 5970 patients were included, of which 855 (14.3%) survived. A three-variable model was determined to be most parsimonious. Prehospital ROSC, age, and initial heart rhythm were identified for inclusion via random forest selection. Finally, SARICA consisted of these 3 variables and ranged from 0 to 10 points, achieving an area under the ROC (AUC) of 0.87 (95% confidence interval: 0.84-0.90) within the testing cohort. CONCLUSION: We developed and internally validated the SARICA score to accurately predict survival of OHCA patients with ROSC at the time of admission. SARICA is clinically practical and developed using an interpretable machine-learning framework. SARICA has unknown generalizability pending external validation studies.

A Theory-Based, Multidisciplinary Approach to Cocreate a Patient-Centric Digital Solution to Enhance Perioperative Health Outcomes Among Colorectal Cancer Patients and Their Family Caregivers: Development and Evaluation Study.

BACKGROUND: Elective colorectal cancer (CRC) surgeries offer enhanced surgical outcomes but demand high self-efficacy in prehabilitation and competency in self-care and disease management postsurgery. Conventional strategies to meet perioperative needs have not been pragmatic, and there remains a pressing need for novel technologies that could improve health outcomes. OBJECTIVE: The aim of this paper was to describe the development of a smartphone-based interactive CRC self-management enhancement psychosocial program (iCanManage) in order to improve health outcomes among patients who undergo elective CRC surgeries and their family caregivers. METHODS: A multidisciplinary international team comprising physicians, specialist nurses, a psychologist, software engineers, academic researchers, cancer survivors, patient ambassadors, and ostomy care medical equipment suppliers was formed to facilitate the development of this patient-centric digital solution. The process occurred in several stages: (1) review of current practice through clinic visits and on-site observations; (2) review of literature and findings from preliminary studies; (3) content development grounded in an underpinning theory; (4) integration of support services; and (5) optimizing user experience through improving interface aesthetics and customization. In our study, 5 participants with CRC performed preliminary assessments on the quality of the developed solution using the 20-item user version of the Mobile App Rating Scale (uMARS), which had good psychometric properties. RESULTS: Based on the collected uMARS data, the smartphone app was rated highly for functionality, aesthetics, information quality, and perceived impact, and moderately for engagement and subjective quality. Several limiting factors such as poor agility in the adoption of digital technology and low eHealth literacy were identified despite efforts to promote engagement and ensure ease of use of the mobile app. To overcome such barriers, additional app-training sessions, an instruction manual, and regular telephone calls will be incorporated into the iCanManage program during the trial period. CONCLUSIONS: This form of multidisciplinary collaboration is advantageous as it can potentially streamline existing care paths and allow the delivery of more holistic care to the CRC population during the perioperative period. Should the program be found to be effective and sustainable, hospitals adopting this digital solution may achieve better resource allocation and reduce overall health care costs in the long run. TRIAL REGISTRATION: ClinicalTrials.gov NCT04159363; https://clinicaltrials.gov/ct2/show/NCT04159363.

Time series classification using a modified LSTM approach from accelerometer-based data: A comparative study for gait cycle detection.

BACKGROUND: Gait event detection (GED) is an important aspect in identifying and interpret a user's gait to assess gait abnormalities and design intelligent assistive devices. RESEARCH QUESTION: There is a need to develop robust GED models that can accurately detect various gait instances in different scenarios and environments. METHODS: This paper presents a novel method of detecting heel strikes (HS) and toe offs (TO) during the user's gait cycle using a modified Long Short-Term Memory (LSTM) networks approach. The method was tested on a database from Movement Analysis in Real-world Environments using Accelerometers (MAREA) (n = 20 healthy subjects) that consisted of walking and running in indoor and outdoor environments with accelerometers positioned on waist, wrist and both ankles. Modifications include oversampling, composite accelerations and optimizing the LSTM network architecture were made. RESULTS: Performance of our modified model was found to be better than six state-of-the-art GED algorithms, with a median F1 score of 0.98 for Heel Strikes and 0.98 for Toe Offs in the scenario of steady walking in an indoor environment, and a median F1 score of 0.94 for Heel Strikes and 0.68 for Toe-offs in the scenario of walking and running in an outdoor environment. SIGNIFICANCE: This paper highlights the potential of the single proposed model to be an alternative to the six GED models in gait detection under various conditions.

Ensemble of Deep Recurrent Neural Networks for Identifying Enhancers via Dinucleotide Physicochemical Properties.

Enhancers are short deoxyribonucleic acid fragments that assume an important part in the genetic process of gene expression. Due to their possibly distant location relative to the gene that is acted upon, the identification of enhancers is difficult. There are many published works focused on identifying enhancers based on their sequence information, however, the resulting performance still requires improvements. Using deep learning methods, this study proposes a model ensemble of classifiers for predicting enhancers based on deep recurrent neural networks. The input features of deep ensemble networks were generated from six types of dinucleotide physicochemical properties, which had outperformed the other features. In summary, our model which used this ensemble approach could identify enhancers with achieved sensitivity of 75.5%, specificity of 76%, accuracy of 75.5%, and MCC of 0.51. For classifying enhancers into strong or weak sequences, our model reached sensitivity of 83.15%, specificity of 45.61%, accuracy of 68.49%, and MCC of 0.312. Compared to the benchmark result, our results had higher performance in term of most measurement metrics. The results showed that deep model ensembles hold the potential for improving on the best results achieved to date using shallow machine learning methods.

Self-Administered Auricular Acupressure Integrated With a Smartphone App for Weight Reduction: Randomized Feasibility Trial.

BACKGROUND: Obesity is a common global health problem and increases the risk of many chronic illnesses. Given the adverse effects of antiobesity agents and bariatric surgeries, the exploration of noninvasive and nonpharmacological complementary methods for weight reduction is warranted. OBJECTIVE: The study aimed to determine whether self-administered auricular acupressure (AA) integrated with a smartphone app was more effective than using AA alone or the controls for weight reduction. METHODS: This study is a 3-arm randomized waitlist-controlled feasibility trial. A total of 59 eligible participants were randomly divided into either group 1 (AA group, n=19), group 2 (AA plus smartphone app, n=19), or group 3 (waitlist control, n=21). A total of 6 reflective zones or acupoints for weight reduction were chosen. The smartphone app could send out daily messages to the subjects to remind them to perform self-pressing on the 6 ear acupoints. A "date picker" of the 8-week treatment course was used to enable the users to input the compliance of pressing and the number of bowel movement daily instead of using the booklet for recordings. The app also served as a reminder for the subjects regarding the dates for returning to the center for acupoint changing and assessments. Treatment was delivered 2 times a week, for 8 weeks. Generalized estimating equations were used to examine the interactions among the groups before and after intervention. RESULTS: Subjects in group 2 expressed that the smartphone app was useful (7.41 out of 10). The most popular features were the daily reminders for performing self-pressing (88%), the ear diagram indicating the locations and functions of the 6 ear points (71%), and ear pressing method demonstrated in the video scripts (47%). Nearly 90% of the participants completed the 8-week intervention, with a high satisfaction toward the overall arrangement (8.37 out of 10). The subjects in group 1 and 2 achieved better therapeutic effects in terms of body weight, body mass index (BMI), waist circumference, and hip circumference and perceived more fullness before meals than the waitlist controls. Although no significant differences in the pairwise comparisons between the 2 groups were detected (P>.05), the decrease in body weight, BMI, body fat, visceral fat rating and leptin level, and increase in adiponectin level were notable in group 2 before and after the intervention. CONCLUSIONS: The high compliance rate and high satisfaction toward the trial arrangement indicate that AA can be used to achieve weight reduction and applied in future large-scale studies. AA integrated with the smartphone app has a more notable effect than using AA alone for weight reduction. Larger sample size should be considered in future trials to determine the causal relationship between treatment and effect. TRIAL REGISTRATION: ClinicalTrials.gov NCT03442712; https://clinicaltrials.gov/ct2/show/NCT03442712 (Archived by WebCite at http://www.webcitation.org/78L2tO8Ql).

Design and Characterization of a Soft Robotic Therapeutic Glove for Rheumatoid Arthritis.

The modeling and experimentation of a pneumatic actuation system for the development of a soft robotic therapeutic glove is proposed in this article for the prevention of finger deformities in rheumatoid arthritis (RA) patients. The Rehabilitative Arthritis Glove (RA-Glove) is a soft robotic glove fitted with two internal inflatable actuators for lateral compression and massage of the fingers and their joints. Two mechanical models to predict the indentation and bending characteristics of the inflatable actuators based on their geometrical parameters will be presented and validated with experimental results. Experimental validation shows that the model was within a standard deviation of the experimental mean for input pressure range of 0 to 2 bars. Evaluation of the RA-Glove was also performed on six healthy human subjects. The stress distribution along the fingers of the subjects using the RA-Glove was also shown to be even and specific to the finger sizes. This article demonstrates the modeling of soft pneumatic actuators and highlights the potential of the RA-Glove as a therapeutic device for the prevention of arthritic deformities of the fingers.

Computational identification of vesicular transport proteins from sequences using deep gated recurrent units architecture.

Protein function prediction is one of the most well-studied topics, attracting attention from countless researchers in the field of computational biology. Implementing deep neural networks that help improve the prediction of protein function, however, is still a major challenge. In this research, we suggested a new strategy that includes gated recurrent units and position-specific scoring matrix profiles to predict vesicular transportation proteins, a biological function of great importance. Although it is difficult to discover its function, our model is able to achieve accuracies of 82.3% and 85.8% in the cross-validation and independent dataset, respectively. We also solve the problem of imbalance in the dataset via tuning class weight in the deep learning model. The results generated showed sensitivity, specificity, MCC, and AUC to have values of 79.2%, 82.9%, 0.52, and 0.861, respectively. Our strategy shows superiority in results on the same dataset against all other state-of-the-art algorithms. In our suggested research, we have suggested a technique for the discovery of more proteins, particularly proteins connected with vesicular transport. In addition, our accomplishment could encourage the use of gated recurrent units architecture in protein function prediction.

Effects of Mattress Material on Body Pressure Profiles in Different Sleeping Postures.

OBJECTIVES: This study compared the body contact pressure profiles of 2 types of mattresses: latex and polyurethane. METHODS: Twenty participants were required to lie down on the different mattresses in 3 different postures for 6 minutes, and their body contact pressure profiles were recorded with a pressure mat sensor. RESULTS: The data indicated that the latex mattress was able to reduce the peak body pressure on the torso and buttocks and achieve a higher proportion of low-pressure regions compared with the polyurethane mattress. CONCLUSIONS: Latex mattress reduced peak body pressure and achieved a more even distribution of pressure compared with polyurethane mattress across different sleeping postures.