2020/12/10 by Jeremy Levy, Levy, Jeremy, Daniel Álvarez +4
Engineering · Medicine · #Chronic Obstructive Pulmonary Disease (COPD) Research #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Non-Invasive Vital Sign Monitoring #Obstructive Sleep Apnea Research #Signal Processing (eess.SP) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2012.05492
openalex publication_date 2020/12/10 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Objective: Chronic obstructive pulmonary disease (COPD) is a highly prevalent\nchronic condition. COPD is a major source of morbidity, mortality and\nhealthcare costs. Spirometry is the gold standard test for a definitive\ndiagnosis and severity grading of COPD. However, a large proportion of\nindividuals with COPD are undiagnosed and untreated. Given the high prevalence\nof COPD and its clinical importance, it is critical to develop new algorithms\nto identify undiagnosed COPD, especially in specific groups at risk, such as\nthose with sleep disorder breathing. To our knowledge, no research has looked\nat the feasibility of COPD diagnosis from the nocturnal oximetry time series.\nApproach: We hypothesize that patients with COPD will exert certain patterns\nand/or dynamics of their overnight oximetry time series that are unique to this\ncondition. We introduce a novel approach to nocturnal COPD diagnosis using 44\noximetry digital biomarkers and 5 demographic features and assess its\nperformance in a population sample at risk of sleep-disordered breathing. A\ntotal of n=350 unique patients polysomnography (PSG) recordings. A random\nforest (RF) classifier is trained using these features and evaluated using the\nnested cross-validation procedure. Significance: Our research makes a number of\nnovel scientific contributions. First, we demonstrated for the first time, the\nfeasibility of COPD diagnosis from nocturnal oximetry time series in a\npopulation sample at risk of sleep disordered breathing. We highlighted what\ndigital oximetry biomarkers best reflect how COPD manifests overnight. The\nresults motivate that overnight single channel oximetry is a valuable pathway\nfor COPD diagnosis.\n