Deep learning approaches for automatic detection of sleep apnea events from an electrocardiogram
رویکردهای یادگیری عمیق برای تشخیص خودکار رویدادهای sleep apnea از الکتروکاردیوگرام-2019
Background and Objective: This study demonstrates deep learning approaches with an aim to find the optimal method to automatically detect sleep apnea (SA) events from an electrocardiogram (ECG) signal. Methods: Six deep learning approaches were designed and implemented for automatic detection of SA events including deep neural network (DNN), one-dimensional (1D) convolutional neural networks (CNN), two-dimensional (2D) CNN, recurrent neural networks (RNN), long short-term memory, and gated- recurrent unit (GRU). Designed deep learning models were analyzed and compared in the performances. The ECG signal was pre-processed, normalized, and segmented into 10 s intervals. Subsequently, the sig- nal was converted into a 2D form for analysis in the 2D CNN model. A dataset collected from 86 patients with SA was used. The training set comprised data from 69 of the patients, while the test set contained data from the remaining 17 patients. Results: The accuracy of the best-performing model was 99.0%, and the 1D CNN and GRU models had 99.0% recall rates. Conclusions: The designed deep learning approaches performed better than those developed and tested in previous studies in terms of detecting SA events, and they could distinguish between apnea and hypopnea events using an ECG signal. The deep learning approaches such as 1D CNN and GRU can be helpful tools to automatically detect SA in sleep apnea screening and related studies.
Keywords: Sleep apnea | Deep learning | Convolutional neural network | Recurrent neural network | Long short-term memory | Gated-recurrent unit
Automatic driver stress level classification using multimodal deep learning
Automatic driver stress level classification using multimodal deep learning-2019
Stress has been identified as one of the contributing factors to vehicle crashes which create a significant cost in terms of loss of life and productivity for governments and societies. Motivated by the need to ad- dress the significant costs of driver stress, it is essential to build a practical system that can detect drivers’ stress levels in real time with high accuracy. A driver stress detection model often requires data from dif- ferent modalities, including ECG signals, vehicle data (e.g., steering wheel, brake pedal) and contextual data (e.g., weather conditions and other ambient factors). Most of the current works use traditional ma- chine learning techniques to fuse multimodal data at different levels (e.g., feature level) to classify drivers’ stress levels. Although traditional multimodal fusion models are beneficial for driver stress detection, they inherently have some critical limitations (e.g., ignore non-linear correlation across modalities) that may hinder the development of a reliable and accurate model. To overcome the limitations of traditional mul- timodal fusion, this paper proposes a framework based on adopting deep learning techniques for driver stress classification captured by multimodal data. Specifically, we propose a multimodal fusion model based on convolutional neural networks (CNN) and long short-term memory (LSTM) to fuse the ECG, vehicle data and contextual data to jointly learn the highly correlated representation across modalities, after learning each modality, with a single deep network. To validate the effectiveness of the proposed model, we perform experiments on our dataset collected using an advanced driving simulator. In this pa- per, we present a multi-modal system based on the adoption of deep learning techniques to improve the performance of driver stress classification. The results show that the proposed model outperforms model built using the traditional machine learning techniques based on handcrafted features (average accuracy: 92.8%, sensitivity: 94.13%, specificity: 97.37% and precision: 95.00%).
Keywords: Deep learning | Driver stress detection | Convolutional neural network | Long short term memory | ECG signal | Vehicle data
A unique feature extraction using MRDWT for automatic classification of abnormal heartbeat from ECG big data with Multilayered Probabilistic Neural Network classifier
استخراج ویژگی منحصر به فرد با استفاده از MRDWT برای طبقه بندی خودکارضربان قلب غیر طبیعی از داده های بزرگ ECG با چند لایه طبقه بندی احتمالی شبکه عصبی-2018
This paper employs a novel adaptive feature extraction techniques of electrocardiogram (ECG) signal for detection of cardiac arrhythmias using multiresolution discrete wavelet transform from ECG big data. In this paper, five types ECG arrhythmias including normal beats have been classified. The MIT-BIH database of 48 patient records is utilized for detection and analysis of cardiac arrhythmias. Proposed feature extraction utilizes Daubechies as wavelet function and extracts 21 feature points which include the QRS complex of ECG signal. The Multilayered Probabilistic Neural Network (MPNN) classifier is pro posed as the best-suited classifier for the proposed feature. Total 1700 ECG betas were tested using MPNN classifier and compared with other three classifiers Back Propagation (BPNN), Multilayered Perceptron (MLP) and Support Vector Machine (SVM). The system efficiency and performance have been evaluated using seven types of evaluation criteria: precision (PR), F-Score, positive predictivity (PP), sensitivity (SE), classification error rate (CER) and specificity (SP). The overall system accuracy, using MPNN technique utilizing the proposed feature, obtained is 99.53% whereas using BPNN, MLP and SVM provide 97.94%, 98.53%, and 99%. The processing time using MPNN classifier is only 3 s which show that the proposed techniques not only very accurate and efficient but also very quick.
Keywords: Signal processing ، Artificial intelligence ، Pattern recognition ، Soft computing ، Wavelet transform
A Novel Adaptive Feature Extraction for Detection of Cardiac Arrhythmias Using Hybrid Technique MRDWT & MPNN Classifier from ECG Big Data
رویکرد استخراج ویژگی تطبیقی برای تشخیص آریتمی های قلب با استفاده از تکنیک ترکیبی MRDWT و MPNN طبقه بندی از داده بزرگ ECG-2018
The efficient automatic detection of cardiac arrhythmia using a hybrid technique from ECG big data has been proposed with novel feature extraction technique using Multiresolution Discrete Wavelet Transform (MRDWT) and Multilayer Probabilistic Neural Network (MPNN) classifier. Big Data of ECG signals have been selected from MIT–BIH arrhythmia database for detection of two types of arrhythmias LBBB (Left Bundle Branch Block) and RBBB (Right Bundle Branch Block). The proposed technique can accurately detect and classify LBBB and RBBB along with normal heartbeat. A novel and hybrid method of detection of cardiac arrhythmia have four main stages: denoising of raw ECG, baseline wander removal, proposed feature extraction, and detection of abnormal heartbeats using MPNN neural classifier. 8600 ECG beats were selected, including 4200 normal and 4400 abnormal beats (2200 LBBB and 2200 RBBB) were utilized for testing the proposed technique. The detection outcome using MPNN was compared with other two neural classifiers: Feed Forward Neural Network (FFNN) and Back Propagation Neural Network (BPNN) classifiers. The accuracy and efficiency of classifiers performance were attained in terms of CER (Classification Error Rate), SP (Specificity), Se (Sensitivity), Pr (Precision), PPr (Positive Predictivity) and F Score. The system performance is achieved with 96.22%, 97.15% and 99.07% overall accuracy using FFNN, BPNN and MPNN. The average percentage of classification error rate (CER) using MPNN classifier is lowest 0.62% whereas FFNN and BPNN show 2.2% and 1. 90% average CER.
Keywords: Big data ، Cardiac arrhythmias ،Biomedical signal processing ، Artificial intelligence ، Machine learning