Evaluation of Machine-Learning Approaches to Estimate Sleep Apnea Severity From At-Home Oximetry Recordings

• Published in: IEEE journal of biomedical and health informatics Vol. 23; no. 2; pp. 882 - 892
• Main Authors: Gutierrez-Tobal, Gonzalo C., Alvarez, Daniel, Crespo, Andrea, del Campo, Felix, Hornero, Roberto
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.03.2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adult; Apnea; At-home oximetry; Bayesian analysis; Biomedical measurement; Blood; Classification; Complexity theory; Data processing; Desaturation; Diagnosis; Diagnostic systems; Discriminant analysis; ensemble learning; Feature extraction; Home Care Services; Humans; Indexes; Learning algorithms; Machine Learning; Mathematical models; Multilayer perceptrons; neural networks; Neural Networks, Computer; Oximetry; Oxygen content; Patients; Regression analysis; Severity of Illness Index; Signal classification; Signal Processing, Computer-Assisted; Sleep; Sleep apnea; sleep apnea severity; Sleep Apnea Syndromes - blood; Sleep Apnea Syndromes - classification; Sleep Apnea Syndromes - diagnosis; Sleep disorders; Statistical analysis
• ISSN: 2168-2194; 2168-2208; 2168-2208
• DOI: 10.1109/JBHI.2018.2823384

Complexity, costs, and waiting list issues demand a simplified alternative for sleep apnea-hypopnea syndrome (SAHS) diagnosis. The blood oxygen saturation signal (SpO 2 ) carries useful information about SAHS and can be easily acquired from overnight oximetry. In this study, SpO 2 single-channel recordings from 320 subjects were obtained at patients' homes and were used to automatically obtain statistical, spectral, nonlinear, and clinical SAHS-related information. Relevant, nonredundant data from these analyses were subsequently used to train and validate four machine-learning methods with the ability to classify SpO 2 signals into one of the four SAHS-severity degrees (no-SAHS, mild, moderate, and severe). All the models trained (linear discriminant analysis, 1-vs-all logistic regression, Bayesian multilayer perceptron, and AdaBoost) outperformed the diagnostic ability of the conventionally used 3% oxygen desaturation index. An AdaBoost model built with linear discriminants as base classifiers reached the highest figures. It achieved 0.479 Cohen's κ in the SAHS severity classification, as well as 92.9%, 87.4%, and 78.7% accuracies in binary classification tasks using increasing severity thresholds (apnea-hypopnea index: 5, 15, and 30 events/hour, respectively). These results suggest that machine-learning can be used along with SpO 2 information acquired at a patients' home to help in SAHS diagnosis simplification .