Biomedical Signal Analysis for Connected Healthcare

BIOMEDICAL SIGNAL ANALYSIS FOR CONNECTED HEALTHCARE
Copyright
Dedication
About the author
Preface
1 . Opportunities for connected healthcare
	1. Introduction
	2. Internet of things
		2.1 Hardware
		2.2 Software
	3. Internet of medical things
		3.1 Remote health monitoring
		3.2 Smartphone application
	4. Wearables for health monitoring
	5. Biomedical signals
		5.1 ECG signal
		5.2 EEG signal
		5.3 EMG signal
		5.4 PPG signal
		5.5 Speech signal
	6. Objectives and organization of the book
	References
2 . Wearables design
	1. Introduction
	2. Wearables survey
		2.1 EEG-based wearable devices
			2.1.1 About EEG signals: properties and acquisition
			2.1.2 Existing technology, drawbacks, and opportunities
			2.1.3 Comparison with clinical EEG data
		2.2 EMG-based wearable devices
			2.2.1 About EMG signals: properties and acquisition
			2.2.2 Existing technology, drawbacks, and opportunities
			2.2.3 Comparison with clinical EMG data
		2.3 ECG-based wearable devices
			2.3.1 About ECG signals: properties and acquisition
			2.3.2 Existing technology, drawbacks, and opportunities
			2.3.3 Comparison with clinical ECG data
		2.4 Other electronic wearables
			2.4.1 Photoplethysmogram
			2.4.2 Auscultation of body sounds
			2.4.3 Motion and gait analysis
	3. Wearables design considerations
		3.1 Signal factors
		3.2 Human factors
		3.3 Environmental factors
		3.4 Medical factors
		3.5 Economic factors
		3.6 Other critical factors
	4. Open hardware design considerations
		4.1 Allocation of hardware design
		4.2 Hardware requirements and methods
			4.2.1 PPG sensor description and bioinstrumentation
			4.2.2 EMG sensor requirements and description
			4.2.3 ECG sensor requirements and description
			4.2.4 Microphone requirements and description
			4.2.5 Motion analysis IMU requirements and description
			4.2.6 Perspectives on wearables hardware design
	5. Textile wearables
	6. Contactless monitoring
	7. Discussions
	References
3 . Biomedical signals and systems
	1. Introduction
	2. Analog to digital conversion
		2.1 Sampling
		2.2 Quantization
			2.2.1 Noise power
			2.2.2 Signal power: Vp2
	3. Linear systems theory
		3.1 Stability and causality
		3.2 Frequency response
	4. Digital filters design
		4.1 Design of FIR filters
		4.2 Design of IIR filters
			4.2.1 Method 1: Pole-zero placement method of IIR filter design
			4.2.2 Method 2: Impulse-invariant method of IIR filter design
			4.2.3 Method 3: Bilinear z-transform method of IIR filter design
				4.2.3.1 BZT method for LPF design
				4.2.3.2 BZT method for HPF design
		4.3 Phase response considerations
		4.4 Homomorphic filtering
	5. Digital filter realization
		5.1 FIR filter realization
		5.2 IIR filter realization
	6. Applications
		6.1 Application 1: Noise filtering techniques
			6.1.1 Synchronized averaging
			6.1.2 Moving average filter
			6.1.3 Savitzky–Golay filter
		6.2 Application 2: Heart rate estimation
	7. Discussion
	References
4 . Adaptive analysis of biomedical signals
	1. Introduction
	2. Adaptive filter design
	3. Adaptive filter algorithms
		3.1 Search method
		3.2 Least mean squares algorithm
	4. Linear prediction
	5. Time series modeling
		5.1 Gain calculation
		5.2 Selection of AR model order
		5.3 Yule–Walker equations
		5.4 Lattice filter
	6. Applications
		6.1 Interference removal in biomedical signals
		6.2 Adaptive segmentation
		6.3 Summary of parametric representation of a biomedical signal
		6.4 Spectral estimation
	7. Discussion
	References
5 . Advanced analysis of biomedical signals
	1. Introduction
		1.1 Evolution of feature extraction methods
	2. Time-domain analysis
	3. Frequency-domain analysis
	4. Joint time-frequency analysis
		4.1 Short-time Fourier Transform
		4.2 Wavelet transform
		4.3 Wigner–Ville Distributions
	5. Signal decomposition analysis
		5.1 Matching pursuits
		5.2 Empirical mode decomposition
	6. Advanced feature extraction and analysis
		6.1 TFD-based feature analysis methods
		6.2 Significance of feature extraction
	7. Sparse analysis and compressive sensing
		7.1 Sparse representations and dictionary learning
		7.2 Compressive sensing
	8. Discussion
	References
6 . Machine learning for biomedical signal analysis
	1. Introduction
	2. Machine learning fundamentals
	3. Types of machine learning models
	4. Challenges with machine learning models
	5. Feature analysis
		5.1 Types of features
		5.2 Feature normalization
		5.3 Feature selection/ranking
	6. Common machine learning techniques
		6.1 Logistic regression
			6.1.1 General concept
			6.1.2 Biomedical signal analysis considerations of logistic regression
		6.2 Linear discriminant analysis
			6.2.1 General concept
			6.2.2 Biomedical signal analysis considerations of LDA
		6.3 Naive Bayes classifier
			6.3.1 General concept
			6.3.2 Biomedical signal analysis considerations of naive Bayes
		6.4 Decision tree
			6.4.1 General concept
			6.4.2 Biomedical signal analysis considerations of decision tree
		6.5 Support vector machine
			6.5.1 General concept
			6.5.2 Kernel methods for nonlinear data
			6.5.3 Biomedical signal analysis considerations of SVM
		6.6 k-nearest neighbor
			6.6.1 General concept
			6.6.2 Biomedical signal analysis considerations of k-NN
		6.7 K-means clustering (unsupervised approach)
			6.7.1 General concept
			6.7.2 Biomedical signal analysis considerations of K-means
		6.8 Ensemble learning
			6.8.1 General concept
			6.8.2 Biomedical signal analysis considerations of ensemble learning
		6.9 Deep learning
			6.9.1 General concept
			6.9.2 Biomedical signal analysis considerations of deep learning
		6.10 Tiny ML
			6.10.1 General concept
			6.10.2 Biomedical signal analysis considerations of tiny ML
	7. Machine learning performance metrics
		7.1 How to measure the success of an ML classifier?
			7.1.1 ROC curve
			7.1.2 What is the right split between training and test datasets?
			7.1.3 How to deal with small datasets?
	8. Fairness and ethics in ML
	9. Summary
	References
7 . Data connectivity and application scenarios
	1. Introduction
	2. Pulse code modulation
	3. Delta modulation
	4. Lossless data compression
		4.1 Huffman code: an example of prefix free code
		4.2 Lempel–Ziv–Welch
	5. Line coding of waveforms
		5.1 Advantages and disadvantages of line coding
	6. Digital modulation
		6.1 Advantages and disadvantages of digital modulation
	7. Telecommunication networks
	8. Wireless technologies
	9. Mobile health
	10. Electronic medical records
		10.1 Advantages and disadvantages of EMR
	11. Personal health record
		11.1 Advantages and disadvantages of PHRs
	12. Interoperability
	13. Health information security and privacy
	14. Human factors and user experiences
	15. Application scenarios
		15.1 Application scenario 1: Smartwatches and wearables for remote health monitoring
		15.2 Application scenario 2: Textile wearables for telemonitoring of vital signs
		15.3 Application scenario 3: Actigraphy for low-cost applications in monitoring sleep and daily activities
			15.3.1 Proposed encoding scheme
			15.3.2 Validation using machine learning
			15.3.3 Signal-encoding results
			15.3.4 Machine learning validation results of encoding process
	16. Summary
	References
Index