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