LKLe Kang

Le Kang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Le Kang has a total of 7 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    PASS: A Novel PPG-based Aortic Stenosis Screening System

    by Le Kang, Ruotong Yang, Yihui Liu, Anfu Zhou, Huadong Ma, Hao Cui

    Aortic stenosis (AS) is the most prevalent valvular heart disease, potentially leading to severe complications such as heart failure, stroke, and sudden death if left untreated. Many patients remain undiagnosed due to the absence of prominent symptoms. Given the serious consequences of AS and the prevalence of undiagnosed asymptomatic patients, there is an urgent need for large-scale screening to enhance early detection and improve patient prognosis. In this work, we propose PASS, a PPG-based Aortic Stenosis Screening System designed for cost-effective, non-invasive, and scalable AS detection. Unlike previous studies that relied on small-scale cohorts or indirect cardiac signals, PASS leverages fingertip PPG to capture the hemodynamic impact of AS in a more accessible manner. To implement PASS, we develop a customized neural network feature extractor capable of performing segmentation, feature extraction, and fusion to capture subtle AS features within the PPG signals. Furthermore, we enhance the classification robustness of PASS by introducing a metric learning-based classifier designed to address the challenge of significant intra-class variation. We evaluate PASS on a real-world clinical dataset comprising 325 subjects, including 158 AS patients confirmed by echocardiography. Our system achieves 91.92% sensitivity and 92.75% specificity, demonstrating strong potential for population-level AS screening using consumer-grade devices.


  • mmTAA: A Contact-Less Thoracoabdominal Asynchrony Measurement System Based on mmWave Sensing

    by Fenglin Zhang, Zhebin Zhang, Le Kang, Anfu Zhou, Huadong Ma

    Thoracoabdominal Asynchrony (TAA) is a key metric in respiration monitoring, which characterizes the non-parallel periodical motion of human's rib cage (RC) and abdomen (AB) during each breath. Long-term measurement of TAA plays a significant role in respiration health tracking. Existing TAA measurement methods including Respiratory Inductive Plethysmography (RIP) and Optoelectronic Plethysmography (OEP) all intrusive to subjects and have certain requirements on operation conditions, which limit their usage to hospital scenario. To address this gap, we propose mmTAA, the first mmWave-based, non-intrusive TAA measurement system ready for ubiquitous usage in daily-life. In mmTAA, we design a Two-stage RC-AB centroid finding module, aiming to identify the most probable location of RC-AB centroid, which can best represent RC and AB in mmWave sensing scenario. Subsequently, we design TAANet, a novel Convolutional Neural Network (CNN)-based architecture with residual modules, tailored for TAA measurement. Meanwhile, in order to address the imbalance of continuous data, we add imbalance information equalizer including feature and label equalizer during network training. We implement mmTAA on a commonly used multi-antenna mmWave radar. We prototype, deploy and evaluate mmTAA on 25 subjects and 25.7h data in total. mmTAA achieves 4.01$^{\circ }$ MAE and 1.56$^{\circ }$ average error, close to OEP method.