Zekai Wang

ORCID: 0009-0007-8764-0375
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Retinal Imaging and Analysis
  • Artificial Intelligence in Healthcare
  • Cardiovascular Health and Risk Factors
  • Machine Learning in Healthcare
  • Ultrasound in Clinical Applications
  • Machine Learning and Data Classification
  • Atrial Fibrillation Management and Outcomes
  • Ultrasonics and Acoustic Wave Propagation
  • Cardiac Arrest and Resuscitation
  • ECG Monitoring and Analysis
  • Infrastructure Resilience and Vulnerability Analysis
  • Electrical Fault Detection and Protection
  • Context-Aware Activity Recognition Systems
  • Text and Document Classification Technologies
  • Anomaly Detection Techniques and Applications
  • Domain Adaptation and Few-Shot Learning
  • Ultrasound and Hyperthermia Applications
  • Smart Grid Security and Resilience
  • Hemodynamic Monitoring and Therapy

Hebei General Hospital
2020-2025

University of Tennessee at Knoxville
2023-2024

Xi'an Jiaotong University
2024

South Dakota School of Mines and Technology
2023

Oklahoma State University
2022-2023

Wuhan University
2022

Weatherford College
2020

Diabetic retinopathy (DR), a microvascular complication of diabetes, is the leading cause vision loss among working-aged adults. However, due to low compliance rate DR screening and expensive medical devices for ophthalmic exams, many patients did not seek proper attention until develops irreversible stages (i.e., loss). Fortunately, widely available electronic health record (EHR) databases provide an unprecedented opportunity develop cost-effective machine-learning tools detection. This...

10.1109/jbhi.2024.3351949 article EN IEEE Journal of Biomedical and Health Informatics 2024-01-09

Task weighting, which assigns weights on the including tasks during training, significantly matters performance of Multi-task Learning (MTL); thus, recently, there has been an explosive interest in it. However, existing task weighting methods assign only based training loss, while ignoring gap between loss and generalization loss. It degenerates MTL's performance. To address this issue, present paper proposes a novel algorithm, automatically via learning-to-learn paradigm, referred to as...

10.18653/v1/2022.findings-acl.271 article EN cc-by Findings of the Association for Computational Linguistics: ACL 2022 2022-01-01

ABSTRACT Combined cardiopulmonary ultrasound (CCPUS) can be applied to bedside differential diagnosis of acute dyspnea. Pulmonary interstitial fibrosis is usually a chronic lung disease. However, some special diseases cause pulmonary with rapid progression. This study included patient unexplained We excluded cardiogenic edema (CPE) by CCPUS and refused hemodialysis. The characteristics the were similar fibrosis. Finally, was diagnosed antiglomerular basement membrane (anti‐GBM) disease...

10.1002/jcu.24072 article EN Journal of Clinical Ultrasound 2025-05-12

This paper proposes a voltage source converter (VSC) -based AC-DC hybrid distribution system (HDS) resilient model to mitigate power outages caused by wildfires.Before wildfire happens, the public-safety shutoff (PSPS) strategy is applied actively cut some vulnerable lines which may easily cause wildfires, and reinforce that are connected critical loads.To load shedding active line disconnection in PSPS strategy, network reconfiguration before occurrence.During restoration period, repair...

10.17775/cseejpes.2022.05090 article EN cc-by-nc-nd CSEE Journal of Power and Energy Systems 2024-01-01

Diabetic retinopathy (DR) is a microvascular complication of diabetes and leading cause vision loss blindness. Screening early detection DR critical but current screening methods rely on eye care experts expensive medical equipment, which are not available in medically underserved communities. The non-image-based, machine-learning approach this study aims to detect the stage using demographics, comorbidities, routine lab results data, widely for diabetic patients. We develop different...

10.1109/case49997.2022.9926605 article EN 2022 IEEE 18th International Conference on Automation Science and Engineering (CASE) 2022-08-20

Recently, Large Language Models (LLMs) have achieved remarkable success using in-context learning (ICL) in the language domain. However, leveraging ICL capabilities within LLMs to directly predict robot actions remains largely unexplored. In this paper, we introduce RoboPrompt, a framework that enables off-the-shelf text-only through without training. Our approach first heuristically identifies keyframes capture important moments from an episode. Next, extract end-effector these as well...

10.48550/arxiv.2410.12782 preprint EN arXiv (Cornell University) 2024-10-16

Cardiac disease is the leading cause of death in US. Accurate heart detection critical importance for timely medical treatment to save patients' lives. Routine use electrocardiogram (ECG) most common method physicians assess electrical activities and detect possible abnormal cardiac conditions. Fully utilizing ECG data reliable depends on developing effective analytical models. In this paper, we propose a two-level hierarchical deep learning framework with Generative Adversarial Network...

10.48550/arxiv.2210.11408 preprint EN other-oa arXiv (Cornell University) 2022-01-01

Abstract Rationale: Fluid resuscitation manages shock effectively. However, is not always caused by hypovolemia; various types of have variable volumetric reactivity. Combined echocardiography and lung ultrasound (LUS) a new technique for assessing volume status pulmonary edema in these patients. We report case unexplained acute circulatory failure kidney injury (AKI) aggravated active fluid resuscitation. used the critical consultation ultrasonic examination (CCUE) protocol evaluation,...

10.1097/md.0000000000023594 article EN cc-by-nc Medicine 2020-12-17

<p>Diabetic retinopathy (DR), caused by damage to the blood vessels in tissue of retina, is a microvascular complication diabetes. DR leading cause vision loss among working-aged adults. However, due low compliance rate screening and expensive medical devices for ophthalmic exams, many patients did not seek proper attention until develops irreversible stages (i.e., loss). Fortunately, widely available electronic health record (EHR) databases provide an unprecedented opportunity develop...

10.36227/techrxiv.22670059.v1 preprint EN cc-by 2023-04-26

<p>Diabetic retinopathy (DR), caused by damage to the blood vessels in tissue of retina, is a microvascular complication diabetes. DR leading cause vision loss among working-aged adults. However, due low compliance rate screening and expensive medical devices for ophthalmic exams, many patients did not seek proper attention until develops irreversible stages (i.e., loss). Fortunately, widely available electronic health record (EHR) databases provide an unprecedented opportunity develop...

10.36227/techrxiv.22670059 preprint EN cc-by 2023-04-26
Coming Soon ...