Book Review
From the text advertising blurb, “This book presents the theoretical basis and applications of biomedical signal analysis and processing” and “This book is aimed at researchers, graduate students in biomedical signal processing, signal processing, electrical engineering, neuroscience, and computer science,” and aims to present “the theoretical basis and applications of biomedical signal analysis and processing.” It is further noted as the first of a series on Biomedical Signal and Image Processing series texts to be published by CRC Press. The text consists of some 14 chapters subdivided into four sections. Two editors are listed, the contributors page names 11 authors (including the two editors), and some 30+ additional authors are named in individual chapters. An overview by this reviewer follows.
Section 1 of the text is titled “Physiological Signal Processing—Challenges” and consists of four chapters. Chapter 1, “Signal Processing for Understanding Physiological Mechanisms for Information Decoding,” is a rapid-fire overview of a human–computer interface diagram, followed by data preprocessing terminology, time domain feature extraction methodology listing, feature dimension reduction, and pattern reduction method overviews. Chapter 2, “Automated Recognition of Alzheimer’s Dementia: A Review of Recent Developments in the Context of Interspeech ADReSS Challenges,” reviews the techniques used by multiple teams in two different years challenges to classify two and three Alzheimer’s dementia diagnosis parameters. With only 14 pages of text, the reader may be best served here by perusal of the five pages of references. Chapter 3 is titled “Electrogastrogram Signal Processing: Techniques and Challenges with Application for Simulator Sickness Assessment” and is a literature review of relevant electrogastrogram data collection methodology and analysis techniques for (hopefully) predictive analysis of driving simulator (and potentially actual driving) induced motion sickness. Chapter 4, “Impact of Cognitive Demand on the Voice Responses of Parkinson’s Disease and Healthy Cohorts,” summarizes the variations of some 15 variables analyzed from voice responses of Parkinson’s patients (with/without levodopa dosing) comparted to normals in testing using the Stroop test. Further work is recommended by these authors.
Section 2 of the text is titled “EEG—ECG Signal Processing” and consists of five chapters. Chapter 5, “Electroencephalography and Epileptic Discharge Identification,” begins with an overview of electroencephalography (source and characteristics) and the special state of epilepsy (11 pages), then reviews (six pages) some current classifier techniques reported in four relevant publications. Five pages of references follow. A 10-page Chapter 6, titled “A Novel End-to-End Secure System for Automatic Classification of Cardiac Arrhythmia,” follows. The “novelty” here is the proposal that transmitted electrocardiogram (“tiny waveform”) data be encrypted prior to transmission, then decrypted prior to denoising and analysis! Eleven authors present Chapter 7 (23 pages), “Machine Learning for Detection and Classification of Motor Imagery in Electroencephalographic Signals.” An overview of their attempts to map evoked imagined motor movements to collected electroencephalographic signals from some nine volunteers. The text is difficult to follow, as references are imbedded in the introductory text and tabular results are difficult to read and interpret. Eight authors (two in common with the prior chapter) present Chapter 8, “Emotion Recognition from Electroencephalographic and Peripheral Physiological Signals Using Artificial Intelligence with Explicit Features,” which begins with a brief review/listing of 10 studies involving extracting emotional state determinations for various data sets, then outlines their experimental work involving extraction of emotional state data for patient data samples including EEG, ECG, respiration rate, galvanic skin resistance, and temperature. Four different approaches to extracting six (happiness, sadness, disgust, etc.) patient determinants are discussed, and tabular results presented. Ten authors (several in common with the prior two chapters) next present Chapter 9, “Identification of Emotion Parameters in Music to Modulate Human Affective States Towards Emotional Biofeedback as a Therapy Support.” A brief literature review of the effect of music on humans and the quantification of music in terms of valence (unpleasant to pleasant) and arousal (calm to excited) in other studies (nine) is followed by the author’s discussion of different methodologies and results (good) for parsing this data reliably from measurements from a given database. These three chapters are written in a similar fashion and make exhaustive use of tabular data presentations.
Section 3 is labeled: “Gait—Balance Signal Processing” and consists of four short chapters. Chapter 10 is titled “Updated ICA Weight Matrix for Lower Limb Myoelectric Classification.” This chapter briefly (eight pages) overviews the inference of surface limb myography relationships with movement in normal versus knee pathology patients during walking, standing, and sitting episodes. Chapter 11, “Cortical Correlates of Unilateral Transfemoral Amputees during a Balance Control Task with Vibrotactile Feedback,” briefly (12 pages) investigates the positive aspects of amputee stump-sensed center of force detection and feedback on patient’s ability to maintain balance. Chapter 12, “Assessing the Impact of Body Mass Index on Gait Symmetry of Able-Bodied Adults Using Pressure-Sensitive Insole,” concludes that there is some gait asymmetry related to increased subject weight, and thus this item needs to be considered in rehabilitation scenarios. Chapter 13, “Analysis of Lower Limb Muscle Activities during Walking and Jogging at Different Speeds,” investigates six leg muscle EMG patters versus four speeds of ambulation slow and fast walking and slow and fast jogging, concluding that slow jogging involves more of the muscle groups at a higher level.
Lastly, Section 4, titled “Wearables—Sensors Signal Processing,” only contains Chapter 14, “Biosensors in Optical Devices for Sensing and Signal Processing Applications,” a nine-page review of the physics of biosensors and a mention of their potential uses in point of care facilities.
This reviewers’ impression: This text is replete with acronyms, some of which are not defined. A tabular listing (appendix, say) would be useful. The content and comprehensiveness of the chapters is highly variable, due to the range of topics (and depth of same) covered by the mixture of authors. Overall, the text seems a collection of overviews, rather than of tutorial or teaching content.
—Reviewed by Paul H. King
Vanderbilt University