A 2D convolutional neural network to detect sleep apnea in children using airflow and oximetry.

Autor: Jiménez-García J; Biomedical Engineering Group, University of Valladolid, Valladolid, Spain; CIBER-BBN, Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain. Electronic address: jorge.jimenez@gib.tel.uva.es., García M; Biomedical Engineering Group, University of Valladolid, Valladolid, Spain; CIBER-BBN, Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain., Gutiérrez-Tobal GC; Biomedical Engineering Group, University of Valladolid, Valladolid, Spain; CIBER-BBN, Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain., Kheirandish-Gozal L; Department of Child Health and Child Health Research Institute, The University of Missouri School of Medicine, Columbia, MO, USA., Vaquerizo-Villar F; Biomedical Engineering Group, University of Valladolid, Valladolid, Spain; CIBER-BBN, Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain., Álvarez D; Biomedical Engineering Group, University of Valladolid, Valladolid, Spain; CIBER-BBN, Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain; Sleep-Ventilation Unit, Pneumology Department, Río Hortega University Hospital, Valladolid, Spain., Del Campo F; Biomedical Engineering Group, University of Valladolid, Valladolid, Spain; CIBER-BBN, Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain; Sleep-Ventilation Unit, Pneumology Department, Río Hortega University Hospital, Valladolid, Spain., Gozal D; Department of Child Health and Child Health Research Institute, The University of Missouri School of Medicine, Columbia, MO, USA; Department of Medical Pharmacology and Physiology, The University of Missouri School of Medicine, Columbia, MO, USA., Hornero R; Biomedical Engineering Group, University of Valladolid, Valladolid, Spain; CIBER-BBN, Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain.
Jazyk: angličtina
Zdroj: Computers in biology and medicine [Comput Biol Med] 2022 Aug; Vol. 147, pp. 105784. Date of Electronic Publication: 2022 Jun 28.
DOI: 10.1016/j.compbiomed.2022.105784
Abstrakt: The gold standard approach to diagnose obstructive sleep apnea (OSA) in children is overnight in-lab polysomnography (PSG), which is labor-intensive for clinicians and onerous to healthcare systems and families. Simplification of PSG should enhance availability and comfort, and reduce complexity and waitlists. Airflow (AF) and oximetry (SpO 2 ) signals summarize most of the information needed to detect apneas and hypopneas, but automatic analysis of these signals using deep-learning algorithms has not been extensively investigated in the pediatric context. The aim of this study was to evaluate a convolutional neural network (CNN) architecture based on these two signals to estimate the severity of pediatric OSA. PSG-derived AF and SpO 2 signals from the Childhood Adenotonsillectomy Trial (CHAT) database (1638 recordings), as well as from a clinical database (974 recordings), were analyzed. A 2D CNN fed with AF and SpO 2 signals was implemented to estimate the number of apneic events, and the total apnea-hypopnea index (AHI) was estimated. A training-validation-test strategy was used to train the CNN, adjust the hyperparameters, and assess the diagnostic ability of the algorithm, respectively. Classification into four OSA severity levels (no OSA, mild, moderate, or severe) reached 4-class accuracy and Cohen's Kappa of 72.55% and 0.6011 in the CHAT test set, and 61.79% and 0.4469 in the clinical dataset, respectively. Binary classification accuracy using AHI cutoffs 1, 5 and 10 events/h ranged between 84.64% and 94.44% in CHAT, and 84.10%-90.26% in the clinical database. The proposed CNN-based architecture achieved high diagnostic ability in two independent databases, outperforming previous approaches that employed SpO 2 signals alone, or other classical feature-engineering approaches. Therefore, analysis of AF and SpO 2 signals using deep learning can be useful to deploy reliable computer-aided diagnostic tools for childhood OSA.
(Copyright © 2022 The Authors. Published by Elsevier Ltd.. All rights reserved.)
Databáze: MEDLINE
Externí odkaz: