Modelling and Analysis of Active Biopotential Signals in Healthcare

Description

This book looks at analysis and modelling of active biopotential signal processing. It emphasises the real-time challenges in biomedical signal processing that occur due to the complex and non-stationary nature of signals in a variety of applications for analysis, classification and identification of different states for improvement of healthcare systems. The main focus of the book is on modelling; acquisition of biomedical signals for different disorders; implementation of methodologies and their impact on different cases; case studies and research directions; automatic identification of related disorders; design and simulation examples; and issues and challenges. Overall, the book addresses the real-time challenges in biomedical signal processing used in a variety of applications such as analysis, classification and identification of different disorders in healthcare systems. It is a valuable guide for all researchers and practitioners who are engaged in studies and research in the area of biomedical signals and their applications. Key Features Modelling and acquisition of biomedical signals for different disordersImplementation of methodologies and their impact on different casesCase studies and research directionsDesign and simulation examples PRELIMS.pdf Preface Acknowledgements Editor biographies Varun Bajaj G R Sinha Contributor list CH001.pdf Chapter 1 Classification of schizophrenia patients through empirical wavelet transformation using electroencephalogram signals 1.1 Introduction 1.2 Methodology 1.2.1 Dataset 1.2.2 Empirical wavelet transform 1.2.3 Feature extraction 1.2.4 Classification techniques 1.2.5 Performance parameters 1.3 Results and discussion 1.4 Conclusion References CH002.pdf Chapter 2 Fuzzy scale invariant feature transform phase locking value and its application to PTSD EEG data 2.1 Introduction 2.2 Method 2.2.1 FSIFT-PLV 2.2.2 Functional connectivity graph indices 2.3 Data 2.3.1 Synthetic data 2.3.2 EEG data 2.4 Results 2.4.1 Synthetic EEG data 2.4.2 Real EEG data 2.5 Conclusion Acknowledgments References CH003.pdf Chapter 3 Weighted complex network based framework for epilepsy detection from EEG signals 3.1 Introduction 3.2 Weighted complex network based framework 3.2.1 Conversion of EEG signals into the WCN 3.2.2 Statistical feature extraction from the WCN 3.2.3 Evaluation of the AWD using classifiers 3.2.4 Evaluation of performance 3.3 Experimental results and discussion 3.3.1 Experimental data 3.3.2 Results and discussion 3.4 Conclusion References CH004.pdf Chapter 4 Epileptic seizure prediction and onset zone localization using intracranial and scalp electroencephalographic and magnetoencephalographic signals 4.1 Epileptic seizure prediction 4.2 Seizure onset zone identification 4.3 Performance indices 4.4 Conclusion and future scope Acknowledgments References CH005.pdf Chapter 5 Automatic drowsiness detection based on variational non-linear chirp mode decomposition using electroencephalogram signals 5.1 Introduction 5.2 Methodology 5.2.1 Dataset 5.2.2 Variational non-linear chirp mode decomposition (VNCMD) 5.2.3 Feature extraction 5.2.4 Classifiers 5.3 Results and discussion 5.4 Conclusion References CH006.pdf Chapter 6 Noise removal and classification of EEG signals using the Fourier decomposition method 6.1 Introduction 6.2 Related work 6.3 Proposed work 6.3.1 Dataset 6.3.2 The Fourier decomposition method 6.4 Classification 6.5 Experimental results and discussion 6.6 Conclusion and proposed future scope References CH007.pdf Chapter 7 Reliable and accurate information extraction from surface electromyographic signals 7.1 Surface electromyography 7.2 Surface EMG applications 7.3 Challenges in sEMG recording 7.4 Detection of atypical signals in HD-sEMG 7.4.1 Feature extraction 7.4.2 Detection methods 7.5 Myoelectric prosthesis control, a hot topic 7.6 Conclusion and future scope Acknowledgments References CH008.pdf Chapter 8 Classification of physical actions from surface EMG signals using the wavelet packet transform and local binary patterns 8.1 Introduction 8.2 Materials and methods 8.2.1 The wavelet transform 8.2.2 The one-dimensional LBP 8.2.3 The support vector machine classifier 8.2.4 The decision tree classifier 8.2.5 The ensemble bagging classifier 8.2.6 The ensemble boosting classifier 8.2.7 The k-nearest neighbor classifier 8.2.8 The linear discriminant classifier 8.3 Experimental work and results 8.4 Conclusion References CH009.pdf Chapter 9 Empirical wavelet transform based classification of surface electromyogram signals for hand movements 9.1 Introduction 9.2 Dataset 9.3 Overview of empirical wavelet transform 9.4 The proposed method 9.4.1 EWT based decomposition 9.4.2 Feature computation 9.4.3 Feature ranking 9.4.4 Classification 9.5 Simulation results 9.6 Discussion 9.7 Conclusion and future scope References CH010.pdf Chapter 10 Analysis of the muscular activity pattern of recurring physical action 10.1 Introduction 10.2 Analytical expressions of joint moments 10.2.1 Data collection and joint moment analysis 10.3 Myoelectric signals during recursive work 10.3.1 The major muscles of the lower extremity 10.3.2 Data collection and subjects 10.3.3 The myoelectrical signal, electrodes and recording 10.3.4 Crosstalk and muscle movement artefacts 10.3.5 The measured electromyogram 10.3.6 Muscle activity pattern 10.4 Joint force estimation 10.4.1 Joint moment pattern and performance measurement 10.5 Conclusions References CH011.pdf Chapter 11 Cloud-based cardiac health monitoring using event-driven ECG processing and ensemble classification techniques 11.1 Introduction 11.2 Background and literature review 11.3 ECG in healthcare 11.4 The proposed approach 11.4.1 Dataset 11.4.2 The event-driven acquisition 11.4.3 The event-driven segmentation 11.4.4 The adaptive rate resampling and denoising 11.4.5 Extraction of features 11.4.6 Machine learning methods 11.5 The performance evaluation measures 11.5.1 Compression ratio 11.5.2 Computational complexity 11.5.3 Classification accuracy 11.6 Experimental results and discussion 11.6.1 Experimental results 11.6.2 Discussion 11.7 Conclusion Acknowledgments References CH012.pdf Chapter 12 Electrocardiogram beat classification using deep convolutional neural network techniques 12.1 Introduction 12.2 Material and methods 12.2.1 The MIT-BIH database 12.2.2 Producing ECG beat images 12.2.3 Convolutional neural networks (CNNs) 12.2.4 Deep transfer learning (DTL) 12.2.5 Support vector machines 12.2.6 Performance metrics 12.3 Experimental work and results 12.4 Discussion 12.5 Conclusion References CH013.pdf Chapter 13 ECG signal watermarking to enhance the security of telecardiology 13.1 Introduction 13.2 Preliminaries 13.3 Prediction error expansion 13.4 Prediction scheme and ECG database 13.4.1 Deep neural network 13.4.2 ECG database 13.5 Training and embedding 13.5.1 Training 13.5.2 Embedding scheme 1 13.5.3 Embedding scheme 2 13.5.4 Embedding scheme 3 13.6 Improved embedding scheme 13.6.1 The effect of ECG abnormalities 13.6.2 Performance on the ECG-ID database 13.7 Conclusion References CH014.pdf Chapter 14 Statistical measures and analysis in electrocardiogram (ECG) signal processing 14.1 Introduction 14.2 The electrocardiogram (ECG) signal and its characteristics 14.2.1 ECG signal generation 14.2.2 ECG signal characteristics 14.3 Statistical measures and analysis 14.4 Statistical analysis in ECG signal processing 14.5 Conclusion References CH015.pdf Chapter 15 The impact of regional atrophy on Alzheimer’s disease and its identification using 3D texture analysis 15.1 Introduction 15.2 Regional atrophy and Alzheimer’s disease 15.3 Related works 15.3.1 VBM based methods 15.3.2 Texture analysis based methods 15.3.3 Shape analysis based methods 15.3.4 Other methods 15.4 Materials and methods 15.4.1 Dataset 15.4.2 The proposed methodology 15.5 Experiments and results 15.5.1 Experiment 1: Voxel as features (VAF) obtained from GM and WM regions 15.5.2 Experiment 2: Volumetric features evaluated on the 3D-DWT sub-bands obtained from all 116 regions 15.5.3 Experiment 3: Volumetric features evaluated on 3D-DWT sub-bands obtained from the top five regions 15.5.4 Experiment 4: Features obtained after applying feature selection on the features of the top five selected regions 15.5.5 Performance comparison with state-of-art methods 15.6 Conclusions Acknowledgments References