Classification of Sleep Apnea Using ECG Signals With Machine Learning Techniques

In this chapter, the authors use various signal processing techniques to analyze and gain insights on how ECG signals for patients suffering from sleep apnea (sleep apnea or obstructive sleep apnea occurs when the muscles that support the soft tissues in the throat, such as tongue and soft palate, relax temporarily) disease vary with respect to a normal person's ECG. The work has three stages: firstly, to identify waves, complexes, morphology in an ECG which reflect the presence of the disease; second, feature extraction techniques to extract features of ECG such as duration of the wave, amplitude distribution, and morphology classes; and third, detailed clustering (unsupervised) algorithm analysis of the extracted features with efficient feature reduction methodologies such as PCA and LDA. Finally, the authors use supervised machine learning algorithms (SVM, naive Bayes classifier, feed forward neural network, and decision tree) to distinguish between ECG signals with sleep apnea and normal ECG signals.

Machine Learning Techniques to Identify and Characterize Sleep Disorders Using Biosignals

Sleep disorders are related to many different diseases, and they could have a significant impact in patients' health, causing an economic impact to the society and to the national health systems. In the United States, according to information from the Center for Disease Control and Prevention, those disorders are affecting 50-70 million in the adult population. Sleep disorders are causing annually around 40,000 deaths due to cardiovascular problems, and they cost the health system more than 16 billion. In other countries, such as in Spain, those disorders affect up to 48% of the adult population. The main objective of this chapter is to review and evaluate the different machine learning techniques utilized by researchers and medical professionals to identify, assess, and characterize sleep disorders. Moreover, some future research directions are proposed considering the evaluated area.

Effect of Delay in EOG Signals for Eye Movement Recognition

Electrooculogram (EOG) signals as a part of human-controlled interface (HCI) is proposed for detecting the relevant information in EOG with and without delay in movement of eyes. The performance of eye movements is studied with the accuracy in identification of information along with single and double blink. The algorithm consists of a simple first order derivative, threshold windowing technique, and pattern recognition. The EOG pattern recognition was studied with time domain features mean value (MV) and ensemble of MV and zero crossing (ZC). The highest average classification accuracy of 85% and 84.4% is obtained from continuous movement of eyes for three classes (L, R, DB and L, R, SB) with two time-domain features. Further, the accuracy of 90% and 88% from two eye movement detection is obtained.

Real-Time EEG Device for Epilepsy Detection Using Wavelet Transform and Support Vector Machine

This chapter gives an effective and efficient technique that can detect epilepsy in real time. It is low cost, low power, and real-time devices that can easily detect epilepsy. Along with EEG device, one can upgrade with GSM module to alert the doctors and parents of patients about its occurrence to prevent a sudden fall, which may cause injury and death. The accuracy of this EEG device depends on the quality of feature extraction technique and classification algorithm. In this chapter, support vector machine (SVM) is used as a classifier. Wavelet transform gives feature extraction, which helps to train data and to detect normal or seizure patients. Discrete wavelet transform (DWT) decomposes the signals into three decomposition levels. In this detection, mean, median, and non-linear parameter entropy were calculated for every sub-band as key parameters. The extracted features are then applied to SVM classifier for the classification. Better accuracy of classification is obtained using wavelet and SVM classifier.

An Investigation Screen for Sleep Disorders Using AI/ML

In this chapter, the authors have discussed a detailed review on sleep, sleep disorders, and their diagnosis. This chapter provides an insight study of sleep, sleep illness characterized by The International Classification of Sleep Disorders (ICSD), factors affecting sleep, and types of sleep based on age group. Artificial intelligence and machine learning algorithms are also applied in recognizing sleep disorders based on EEG signal attributes. It also highlights the classification of insomnia using different classifiers such as support vector machine, decision tree, and deep neural network.

Medical Signal Processing

This chapter introduces sleep, the pattern of sleep, wakefulness, disorders associated with sleep, diseases of heart and lungs that can be identified by analysing one's sleep. Sleep is generally equated to the neurological system and the brain. It is believed that sleep can be identified only with EEG. This chapter also explores the usage of EEG in detecting the disorders associated with sleep, and more emphasis is given to the bio signals other than EEG, which includes ECG, PPG, acoustic signals that can be used in understanding the sleep and its related disorders. It explains the biomedical devices that are used for sleep-related studies. This chapter explores the stages of sleep signal processing where the authors have suggested how to reduce noises at the stage of data acquisition. Further topics explore various signal processing methods that need to be adapted in various stages, namely preprocessing, filtering, feature extraction, validation, and automated processing.

An Optimized FPGA Architecture for Sleep Level Categorization Using Fuzzed Utilized Machine Learning Approach

In this chapter, an efficient FPGA architecture is proposed to categorize and analyze the sleep level. This proposed architecture is implemented using four sub parts which are namely preprocessing unit, FIR filtering, self-regulated learning, and fuzzy deduction. The EEG (electro encephalo gram) and EMG (electro myogram) are signal samples are considered for the analysis of this sleep level. The signals are initially preprocessed to remove undesired signal components. Further, a reconfigurable multichannel multiply accumulate (MAC)-based FIR filter is utilized for achieving the desired signal. Then the signal is classified based on the reference data with the use of self-regulated machine learning and fuzzy deduction schemes which involves averaging and thresholding process. Further, the signals are categorized into completely awake level, partially awake level, and sleep level using fuzzy if-then rules. The performance parameters are analyzed in terms of sensitivity, specificity, latency, area occupied, power consumption, and speed enhancement.

Obstructive Sleep Apnea Analysis Techniques

Obstructive sleep apnea (OSA) is a common disease in adults between 20 and 100 years of age and its prevalence has been reported to be higher in males than in females. There are several methods for measuring the severity of OSA. These include measuring the number of apnea and hypopneas per hour of sleep (apnea-hypopnea index: AHI), the degree of oxygen desaturation during sleep, or the severity of daytime sleepiness, which is the most common condition that negatively affects the quality of life. The gold standard test for the diagnosis of OSA is polysomnography. The pulse oximeter, home sleep apnea testing, SleepQuest device, peripheral arterial tonometry, ApneaLink Plus device, maximum static inspiratory pressure, and maximum static expiratory pressure measurements are preferred methods for OSA diagnosis.

Contemporary Concepts in the Diagnosis and Management of Obstructive Sleep Apnea

Optimal sleep is an inseparable component of both physical and psychological well-being. With the widespread increase in the prevalence of sleep disorders, there has been an immense interest among the global researchers in exploring the molecular biology of sleep and innovative modalities for diagnosing and treating sleep disorders. Notably, sleep disorders encompass a wide spectrum of sleep disturbances with potential multisystem complications. Polysomnography is an overnight sleep study that is widely considered as the gold standard objective diagnostic method for diagnosing obstructive sleep apnea (OSA) and provision of continuous positive airway pressure (CPAP), which maintains airway patency during sleep remains the cornerstone therapy for OSA. Although CPAP remains the mainstay of therapy for OSA, various oral appliances and surgical interventions have widely been considered for OSA. This chapter provides a comprehensive overview of contemporary diagnostic and therapeutic approaches available in clinical practice for sleep-related breathing disorders with particular emphasis on OSA.

A Review on Sleep Disorder Analysis and Applications Based on Artificial Intelligence

The recent developments in multi-modular sensor technology for analyzing physiological and psychological activities have brought a breakthrough in the biomedical engineering thereby aiding the real-time monitoring of various vital signals. The monitoring of circadian rhythms and EEG signals keeps a proper observation on sleep and analyzes various sleep disorders. The sleep patterns related to disease and wellness applications can be analyzed with the help of multi-sensor-generated data. Several challenges such as performance evaluation, data storage, processing and integration, modeling, and interpretation as well as curation are to be overcome in this field for expansion of this technology in the future. The digitalization of sleep is an interdisciplinary field of research incorporating neuroscience, bioengineering, epidemiology, clinical medicine, computer science, and electrical engineering. This chapter discusses various sleep disorders, AI, analysis, and applications available. Finally, the challenges and future scope are also discussed followed by the conclusion.