Advanced Methods in Biomedical Signal Processing and Analysis

Front Matter
Copyright
Contributors
Feature engineering methods
	Machine learning projects development standards and feature engineering
	Exploratory data analysis
		Types of input data
		Data preparation and preprocessing
			Missing values treatment
			Encoding the categorical variables
			Investigation of the data distribution
			Binning
			Identifying and treatment of outliers
			Variable transformation
			Min-max scaling
			Logarithm transformation
			Centering and scaling
			Box-Cox normalization
	Data vs features
		Relations between data and features
		Feature extraction methods
			Linear vs nonlinear
			Multivariate vs univariate
		Curse of dimensionality
			Data sparsity
			Distance concentration
			Avoiding the curse of dimensionality
	Feature reduction
	Feature selection
		Unsupervised feature selection
		Supervised feature selection
			Exhaustive search
			Filter methods
			Wrapper methods
			Embedded methods
	Feature dimensionality reduction
		Principal component analysis
		Independent component analysis
		Nonnegative matrix factorization
		Self-organizing maps
		Autoencoders
	Concluding remarks
	References
Heart rate variability
	Introduction
		Effects of blood pressure on HRV
		Effect of myocardial infarction on HRV
		Relation between HRV and cardiac arrhythmia
		Relation between HRV, age, and gender
		Effects of drugs, alcohol, and smoking on HRV parameters
		Effects of menstrual cycle on HRV parameters
	Literature review
		Analyses of time-domain parameters
		Analysis of frequency-domain HRV parameters
		Classification and prediction of ECG signals
	Results
		Statistical analysis
		Machine learning results
		Variation of HRV during the menstrual cycle
	Discussion
	Conclusion
	Acknowledgments
	References
Understanding the suitability of parametric modeling techniques in detecting the changes in the HRV signals ac ...
	Introduction
	Literature review on cannabis and its legal status
	Methods
		Acquisition of the ECG signals and extraction of the HRV signals
		Parametric modeling of the HRV signals
		Statistical analysis
		Development of ML classifiers
			Selection of input parameters
			Machine learning techniques
	Results
		AR modeling of the HRV signals
		MA modeling of the HRV signals
		ARMA modeling of the HRV signals
		Development of ML-based classifiers using the coefficients of all the parametric models of the HRV signals
	Discussion
	Conclusion
	Conflict of interest statement
	References
Patient-specific ECG beat classification using EMD and deep learning-based technique
	Introduction
	Database
	Proposed methodology
		Preprocessing
			Noise removal using EMD technique
		Deep learning-based architecture for ECG beat classification
	Experimental results
		Performance metrics
		Selection of hyperparameters for the proposed model
		Performance of the proposed system for ECG beat detection
		Comparison of the proposed framework with state-of-the-art techniques
	Conclusions
	References
Empirical wavelet transform and deep learning-based technique for ECG beat classification
	Introduction
		Related works and motivation
	Database
	Proposed methodology
		Preprocessing
		Deep learning architecture for ECG beat classification
	Experimental results
		Preprocessing of ECG beats using EWT technique
		Metrics utilized to assess the performance of the EWT-based deep learning technique
		Parameters optimization of the deep learning-based model
		Performance of the proposed EWT-based deep learning classifier
		Performance comparison of the proposed EWT-based deep learning technique with state-of-the-art techniques
	Conclusions
	References
Development of an internet of things (IoT)-based pill monitoring device for geriatric patients
	Introduction
	Literature review
	Materials and methods
		Materials and softwares
		Methods
			Designing the medication monitoring system
				Designing the hardware component
				Development of the software for medication monitoring
	Results and discussions
		Developing the medication monitoring system
		Testing the medication monitoring system
		Discussions
	Conclusion
	Conflict of interest
	Appendix
	References
Chapter 7: Biomedical robotics
	1. Introduction
	2. Challenges and opportunities
	References
Combating COVID-19 by employing machine learning predictions and projections
	Introduction
	COVID-19: The 2020 pandemic
		Origin and classification
		The genome
		Epidemiology
			Source and spectrum of infection
			Disease etiology
			Pathogenesis
		Treatments
	What is machine learning (ML)?
		What does ML do?
		What is data in ML?
		Framework of ML-based prediction and projections
		Demystifying machine learning
		Machine learning: The process
		Types of machine learning
			Supervised learning
			Unsupervised learning
			Semi-supervised learning
			Reinforcement learning
	Key application of machine learning with illustrative examples: Fighting COVID-19
		Pandemic preparedness
		Risk assessment and priority testing
		Digital contact tracing
		Integrated diagnosis
		Assisting drug discovery process
		Aiding in vaccine development
	Concerns
	Final thoughts
	Takeaway points
	References
Deep learning methods for analysis of neural signals: From conventional neural network to graph neural network
	Introduction
	Deep learning methods
		Some discussion on CNNs and RNNs
		Hybrid models
		Attention mechanism in deep learning
	Graph neural network
		Transition from basic models to graph-based models
		GNN: Convolutional, attention, and message passing flavors
		Dynamic GNNs
	Applications of GNNs on neural data
	Discussion
	References
Improved extraction of the extreme thermal regions of breast IR images
	Introduction
	Methodology
		Graph theory
	Experimental results and discussion
		Case 1: Breast cancer
		Case 2: Breast cancer
		Case 3: Mammary cysts
		Case 4: Mammary cysts
		Case 5: Benign tumor
		Case 6: Benign tumors (multiple)
		Case 7: Benign tumors
		Case 8: Benign tumors
		Case 9: Advanced cancer
	Conclusion
	Acknowledgments
	References
New metrics to assess the subtle changes of the heart's electromagnetic field
	Introduction
	Information technology of magnetocardiography: Basis, technical means, diagnostic metrics
		Definition and short essay on the history of magnetocardiography
		Technical means: Magnetometric complex
		The consecutive steps, electrophysiological basis, and algorithms of the MCG-signal analysis
		Inverse problem statement
		Algorithm for solving the inverse problem of magnetostatics for a 2D field source
		Consideration of the spatial configuration of the magnetic flux transformer: Axial and planar gradiometers
		Application of the algorithm for the analysis of the magnetocardiosignal
	Metrics and information technologies for the analysis of magnetocardiographic data based on two-dimensional visualizat ...
		Clinical approbation of metrics of analysis of magnetocardiographic data based on two-dimensional visualization of t ...
		Metrics and information technologies of analysis of magnetocardiographic data based on three-dimensional visualizati ...
	New metrics and information technologies based on computerized electrocardiography
		Principles of the electrocardiogram-scaling technique for detecting subtle changes
		Clinical approbation of new information technologies and metrics of computerized electrocardiography
	New metrics and information technologies based on heart rate variability analysis
		Heart rate variability and pain analysis
	Conclusions
	Acknowledgments
	References
	Further reading
The role of optimal and modified lead systems in electrocardiogram
	Introduction
		Lead theory
		BSPM
		Modifications in standard ECG lead system
	Modified and optimal lead systems
		Bipolar monitoring leads
		Modified chest leads
		Minimal monitoring leads
		Mason-Likar lead system
		Lund lead system
		Derived 12-lead systems
		Lewis lead
		Modified Lewis lead
		EASI lead system
		Fontaine bipolar leads
		Modified limb lead system
		Monitoring neonatal and pediatric ECG
		P-lead system
	ECG signal processing
		Data acquisition
		Denoising techniques
		Feature extraction techniques
		Signal processing techniques
		Classification techniques
	Advantages of optimal and modified leads in ECG signal processing
	Conclusion
	Acknowledgments
	References
Adaptive rate EEG processing and machine learning-based efficient recognition of epilepsy
	Introduction
	The electroencephalogram (EEG) for healthcare
	Signal acquisition, preprocessing, and features extraction
		Dataset
		Reconstruction
		Adaptive rate acquisition
		Adaptive rate segmentation
		Adaptive rate interpolation
		Adaptive rate filtering
		Feature extraction method
		Machine learning methods
			K-nearest neighbors (K-NN)
			Artificial neural network (ANN)
			Support vector machines (SVM)
		The performance evaluation measures
			Samples ratio
			Compression ratio
			Accuracy (ACC)
			Specificity (SP)
			F-measure (F1)
			Kappa index (kappa)
	Results
	Discussion and conclusion
	Acknowledgments
	References
Development of a novel low-cost multimodal microscope for food and biological applications
	Introduction
	Literature review
	Materials and methods
		Material and software
		Development of the microscope
			The sample magnification and imaging assembly (SMIA)
			The sample illuminator assembly (SIA)
			The filter holder assembly (FHA)
			The focus adjustment assembly (FAA)
			The sample stage movement assembly (SSMA)
		Development of the optical filter set of the microscope
		Development of the software
		Testing of the developed microscope
			Microbes and milk protein
			Melted chocolate
			Wastewater cultured microbes
			Oleogels
	Results and discussion
		Development of the microscope
		Testing of the developed microscope
	Conclusion and future scope
	Acknowledgments
	References
Index
	A
	B
	C
	D
	E
	F
	G
	H
	I
	J
	K
	L
	M
	N
	O
	P
	R
	S
	T
	U
	V
	W
	X
	Y
	Z