A5052854031- ECG Monitoring and Analysis
- EEG and Brain-Computer Interfaces
- Non-Invasive Vital Sign Monitoring
- Analog and Mixed-Signal Circuit Design
- Phonocardiography and Auscultation Techniques
- Blind Source Separation Techniques
- Cardiac electrophysiology and arrhythmias
- Fault Detection and Control Systems
- Ferroelectric and Negative Capacitance Devices
- Advanced Combustion Engine Technologies
- Rough Sets and Fuzzy Logic
- Stock Market Forecasting Methods
- Forecasting Techniques and Applications
- Advanced Memory and Neural Computing
- Low-power high-performance VLSI design
- Hydraulic and Pneumatic Systems
- Energy Load and Power Forecasting
- Text and Document Classification Technologies
- Engineering Diagnostics and Reliability
- Chronic Obstructive Pulmonary Disease (COPD) Research
- Radiomics and Machine Learning in Medical Imaging
- Advancements in PLL and VCO Technologies
- Lung Cancer Diagnosis and Treatment
- Engineering Applied Research
- Machine Fault Diagnosis Techniques
North University of China
2021-2022
China Jiliang University
2020-2021
Alibaba Group (United States)
2019-2021
Alibaba Group (Cayman Islands)
2020
QRS detection is a crucial step in analyzing the electrocardiogram (ECG). For ECG collected by wearable devices, robust algorithm that yields high accuracy spite of abnormal morphologies and severe noise needed. In this paper, we propose method based on high-resolution wavelet packet decomposition (HR-WPD) convolutional neural network (CNN). Firstly, design HR-WPD decomposes into multiple signals with different frequency bands to provide detailed features. Secondly, all decomposed are...
Long-term Electrocardiogram (ECG) analysis has become a common means of diagnosing cardiovascular diseases. In order to reduce the workload cardiologists and accelerate diagnosis, an automated patient-specific heartbeat classification method based on customized convolutional neural network (CNN) is proposed in this paper. The parallel layers with kernels different receptive fields are responsible for extracting multi-spatial deep features heartbeats, channel-wise attention module adopted...
This paper presents a neural network based processor with improved computation efficiency, which aims at multiclass heartbeat recognition in wearable devices. A lightweight classification algorithm that integrates both bi-directional long short-term memory (BLSTM) and convolutional networks (CNN) is proposed to deliver high accuracy minimal scale. To reduce energy consumption of the algorithm, similarity between consecutive heartbeats exploited achieve degree reuse hardware architecture. In...
Long-term electrocardiogram (ECG) monitoring requires high-ratio lossless compression techniques to reduce data transmission energy and storage capacity. In this paper, we have proposed a ECG system with low computational complexity. Firstly, as the morphologies of change over time, divide signal each heartbeat cycle into two regions. To achieve high prediction accuracy, 1 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">st</sup> order linear...
In the fault diagnosis of high-pressure common rail diesel engines, it is often necessary to face problem insufficient diagnostic training samples due high cost obtaining or difficulty samples, resulting in inability diagnose state. To solve above problem, this paper proposes a small-sample method for system using learning based on data augmentation and GA_BP neural network. The synthesis set Least Squares Generative Adversarial Networks (LSGANs) improves quality diversity synthesized data....
This paper presents an automated patient-specific ECG classification algorithm, which integrates long short-term memory (LSTM) and convolutional neural networks (CNN). While LSTM extracts the temporal features, such as heart rate variance (HRV) beat-to-beat correlation from sequential heartbeats, CNN captures detailed morphological characteristics of current heartbeat. To further improve performance, adaptive segmentation re-sampling are applied to align heartbeats different patients with...
Recent works show that multi-bit multiplications can be achieved in multi-cycles via serial in/near memory computing, aiming at reducing data transfers hence improving energy efficiency. However, the pure approach suffers from a long latency and leaves additional room to optimize We propose three new techniques develop novel SRAM structure realize multiplication with bi-serial computing. Firstly, we use 2-bit binary numbers (bi-serial) as smallest unit of operation working mixed-signal mode...
In the presence of premature atrial contraction (PAC), ventricular (PVC) or other ectopic beats, RR intervals (RRIs) may be disturbed, which results in types heart disease being misdiagnosed as fibrillation (AF). this study, a low-complexity AF detection method based on short ECG is proposed, includes RRIs modification and feature selection. The extracted are used to determine whether potential RRI interference exists modify it. Next, modified RRIs, features evaluated selected by methods...
The performance of the diesel engine was determined by injection pressure high common rail system. observation method curve great value in evaluating operating conditions engine. However, could make it impossible to directly observe data on a system operation. In this paper, signals and control pulse widths, which were easy collect, used as basis for proposing GRU (Gated Recurrent Unit) based model curves order solve above problem. accuracy precision verified experiments compared with...
Background: To evaluate the performance of a machine learning approach to predict pulmonary function test (PFT) result from parameter response mapping (PRM), which enhances value one-stop CT scanning.Methods: 615 subjects with PFT and paired inspiratory expiratory chest scanning were enrolled retrospectively, classified into normal group, high risk group COPD based on PFT. 72 PRM-derived quantitative parameters including volume (cc) percentage (%) emphysema, functional-small airways disease...