A5060704848- Machine Learning in Healthcare
- Cardiac, Anesthesia and Surgical Outcomes
- Cardiac Arrest and Resuscitation
- Artificial Intelligence in Healthcare and Education
- Mechanical Circulatory Support Devices
- Winter Sports Injuries and Performance
- Sports Dynamics and Biomechanics
- Martial Arts: Techniques, Psychology, and Education
- Sepsis Diagnosis and Treatment
- Phonocardiography and Auscultation Techniques
- Heart Failure Treatment and Management
- COVID-19 diagnosis using AI
- Family and Disability Support Research
- Clinical Reasoning and Diagnostic Skills
- Simulation and Modeling Applications
- Cerebral Palsy and Movement Disorders
- Music and Audio Processing
- Wireless Sensor Networks and IoT
- Respiratory Support and Mechanisms
- Infant Development and Preterm Care
- Advanced Decision-Making Techniques
- Advanced Computational Techniques and Applications
- Respiratory and Cough-Related Research
Washington University in St. Louis
2022-2024
Shenzhen Blood Center
2009
Henan University of Urban Construction
2007
Pingdingshan University
2007
General Administration of Sport of China
2002
Sleep issues are common in children with cerebral palsy (CP), although there challenges obtaining objective data about their sleep patterns. Actigraphs measure movement to quantify but accuracy CP is unknown. Our goals were validate actigraphy for assessment and study patterns a cross-sectional cohort study.We recruited (N = 13) without aged 2-17 years (mean age 9 y 11mo [SD 4 10mo] range 4-17 y; 17 males, females; 54% spastic quadriplegic, 23% diplegic, 15% hemiplegic, 8% unclassified CP)....
Smart wearable sensors are essential for continuous health-monitoring applications and detection accuracy of symptoms energy efficiency processing algorithms key challenges such devices. While several machine-learning-based the abnormal breath sounds reported in literature, they either too computationally expensive to implement into a device or inaccurate multi-class detection. In this paper, kernel-like minimum distance classifier (K-MDC) acoustic signal devices was proposed. The proposed...
Given the risks and cost of hospitalization, there has been significant interest in exploiting machine learning models to improve perioperative care. However, due high dimensionality noisiness data, it remains a challenge develop accurate robust encoding for surgical predictions. Furthermore, is important be interpretable by care practitioners facilitate their decision making process. We proposeclinical variational autoencoder (cVAE), deep latent variable model that addresses challenges...
Abstract Background Veno‐venous extracorporeal membrane oxygenation (V‐V ECMO) is a lifesaving support modality for severe respiratory failure, but its resource‐intensive nature led to significant controversy surrounding use during the COVID‐19 pandemic. We report performance of several ECMO mortality prediction and severity illness scores at discriminating survival in large V‐V cohort. Methods validated ECMOnet, PRESET (PREdiction Survival on Therapy‐Score), Roch, SOFA (Sequential Organ...
Abstract Objective Extracorporeal membrane oxygenation (ECMO) resource allocation tools are currently lacking. We developed machine learning (ML) models for predicting COVID-19 patients at risk of receiving ECMO to guide patient triage and allocation. Material Methods included admitted intensive care units >24 h from March 2020 October 2021, divided into training testing development testing-only holdout cohorts. deployment timely prediction model ForecastECMO using Gradient Boosting...
Extracorporeal membrane oxygenation (ECMO) is an essential life-supporting modality for COVID-19 patients who are refractory to conventional therapies. However, the proper treatment decision has been subject of significant debate and it remains controversial about benefits from this scarcely available technically complex option. To support clinical decisions, a critical need predict potential no-treatment responses. Targeting challenge, we propose Treatment Variational AutoEncoder (TVAE),...
Early detection of surgical complications allows for timely therapy and proactive risk mitigation. Machine learning (ML) can be leveraged to identify predict patient risks postoperative complications. We developed validated the effectiveness predicting using a novel Variational Autoencoder (surgVAE) that uncovers intrinsic patterns via cross-task cross-cohort presentation learning. This retrospective cohort study used data from electronic health records adult patients over four years (2018 -...
Abstract Objective Early detection of surgical complications allows for timely therapy and proactive risk mitigation. Machine learning (ML) can be leveraged to identify predict patient risks postoperative complications. We developed validated the effectiveness predicting using a novel Variational Autoencoder (surgVAE) that uncovers intrinsic patterns via cross-task cross-cohort presentation learning. Materials Methods This retrospective cohort study used data from electronic health records...
Abstract Objective Extracorporeal membrane oxygenation (ECMO) is among the most resource-intensive therapies in critical care. The COVID-19 pandemic highlighted lack of ECMO resource allocation tools. We aimed to develop a continuous risk prediction model enhance patient triage and allocation. Material Methods leveraged multimodal data from National COVID Cohort Collaborative (N3C) hierarchical deep learning model, labeled “PreEMPT-ECMO” (Prediction, Early Monitoring, Proactive Triage for...
Major postoperative complications are devastating to surgical patients. Some of these potentially preventable via early predictions based on intraoperative data. However, data comprise long and fine-grained multivariate time series, prohibiting the effective learning accurate models. The large gaps associated with clinical events protocols usually ignored. Moreover, deep models generally lack transparency. Nevertheless, interpretability is crucial assist clinicians in planning for delivering...
Major postoperative complications are devastating to surgical patients. Some of these potentially preventable via early predictions based on intraoperative data. However, data comprise long and fine-grained multivariate time series, prohibiting the effective learning accurate models. The large gaps associated with clinical events protocols usually ignored. Moreover, deep models generally lack transparency. Nevertheless, interpretability is crucial assist clinicians in planning for delivering...