ZMZhen Mei

Zhen Mei

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Zhen Mei has a total of 21 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • A PPG sensor-based multi-feature machine learning approach for wide-range and high-precision SpO 2 measurement

    by Lian Guo, Zhen Chen, Yijun Li, Haiyang Hu, De He, Zhen Mei, Xuesheng Ma

    Objective . This study aimed to improve peripheral oxygen saturation ( SpO 2 ) estimation across a wide range for hypoxemic-risk assessment in chronic obstructive pulmonary disease (COPD) using dual-wavelength photoplethysmography (PPG) and machine learning (ML). Approach . A MAX30102-based module acquired synchronized red (660 nm) and near-infrared (880 nm) PPG signals from the right index finger. A CMS50DA oximeter on the right middle finger provided concurrent reference SpO 2 readings for training and evaluation. Smartphone video captured its display for manual label extraction and independent checking. The ratio-of-ratios ( R ) and morphological, temporal, statistical, and derivative features were extracted. Principal component analysis-based feature selection identified a feature subset. ML and PPG-oriented deep-learning models were compared with conventional R -ratio calibration using mean absolute error (MAE), root mean squared error (RMSE), and the coefficient of determination ( R 2 ). Main results . The Extra Trees regressor achieved the lowest RMSE among evaluated models, with an MAE of 0.38 percentage points, an RMSE of 0.90 percentage points, and an R 2 of 0.91. It outperformed conventional R -ratio calibration. In the hypoxemic range ( 70 % ⩽ SpO 2 ⩽ 90 % ), MAE was 0.98 percentage points, RMSE was 2.04 percentage points, and R 2 was 0.66. Significance . Combining dual-wavelength PPG with multi-domain features and ML shows potential for wide-range SpO 2 estimation and hypoxemic-risk assessment in COPD.

  • A High-Dynamic-Range SpO₂ Measurement Method Using a Smartphone Camera

    by Lu Luo, Tianju Li, Kai Du, Zhiyi Huang, Veronica Liesaputra, Zhen Mei, Yuan-Ting Zhang, Xuesheng Ma

    The narrow range of blood oxygen saturation (SpO2) measurement overlooks severe hypoxemia patients and some mild hypoxemia patients, preventing them from receiving accurate and timely warnings, which increases the risk of disease deterioration and related complications. In this paper, we present an SpO2 measurement method from fingertip videos, covering an estimation range from 70% to 100%. Our approach incorporates the Transformer Encoder mixed with Squeeze and Excitation Block (SETE) network architecture and the signal quality assessment module with filtering function. The SETE model itself features powerful global information sensing capability and adopts a channel auto-adjustment weighting mechanism to promptly capture dynamic changes in SpO2 levels. The public dataset used in this work maintains subjects’ SpO2 levels over a wide range by controlling for the percentage of oxygen inhaled by participants, thus keeping low SpO2 values over an extended duration. Comparative tests demonstrate that the SETE model outperforms current state-of-the-art contact SpO2 measurements in terms of performance. Experimental results indicate the mean absolute error (MAE) of 1.35% and the root mean square error (RMSE) of 2.08% between the estimated and reference SpO2 values. Consequently, this method offers a novel and effective approach for users to conduct SpO2 measurements based on electronic devices.


  • A Non-Invasive Blood Glucose Detection System Based on Photoplethysmogram With Multiple Near-Infrared Sensors

    by Kai Du, Jing Li, Zhiyi Huang, Houbing Song, Yuanting Zhang, Yuan Zhang, Zhen Mei

    Recent advancements in non-invasive blood glucose detection have seen progress in both photoplethysmogram and multiple near-infrared methods. While the former shows better predictability of baseline glucose levels, it lacks sensitivity to daily fluctuations. Near-infrared methods respond well to short-term changes but face challenges due to individual and environmental factors. To address this, we developed a novel fingertip blood glucose detection system combining both methods. Using multiple light sensors and a lightweight deep learning model, our system achieved promising results in oral glucose tolerance tests. A total of 10 participants were involved in the study, each providing approximately 700 data segments of about 10 seconds each. With a root mean squared error of 0.242 mmol/L and 100% accuracy in zone A of the Parkes error grid, our approach demonstrates the potential of multiple near-infrared sensors for non-invasive glucose detection.


  • ECGAN-Assisted ResT-Net Based on Fuzziness for OSA Detection

    by Zhiya Wang, Xue Pan, Zhen Mei, Zhifei Xu, Yudan Lv, Yuan Zhang, Cuntai Guan

    Objective: Growing attention has been paid recently to electrocardiogram (ECG) based obstructive sleep apnea (OSA) detection, with some progresses been made on this topic. However, the lack of data, low data quality, and incomplete data labeling hinder the application of deep learning to OSA detection, which in turn affects the overall generalization capacity of the network. Methods: To address these issues, we propose the ResT-ECGAN framework. It uses a one-dimensional generative adversarial network (ECGAN) for sample generation, and integrates it into ResT-Net for OSA detection. ECGAN filters the generated ECG signals by incorporating the concept of fuzziness, effectively increasing the amount of high-quality data. ResT-Net not only alleviates the problems caused by deepening the network but also utilizes multi-head attention mechanisms to parallelize sequence processing and extract more valuable OSA detection features by leveraging contextual information. Results: Through extensive experiments, we verify that ECGAN can effectively improve the OSA detection performance of ResT-Net. Using only ResT-Net for detection, the accuracy on the Apnea-ECG and private databases is 0.885 and 0.837, respectively. By adding ECGAN-generated data augmentation, the accuracy is increased to 0.893 and 0.848, respectively. Conclusion and significance: Comparing with the state-of-the-art deep learning methods, our method outperforms them in terms of accuracy. This study provides a new approach and solution to improve OSA detection in situations with limited labeled samples.