Sleep apnea and respiratory anomaly detection from a wearable band and oxygen saturation

Abstract

Objective: Sleep related respiratory abnormalities are typically detected using polysomnography. There is a need in general medicine and critical care for a more convenient method to automatically detect sleep apnea from a simple, easy-to-wear device. The objective is to automatically detect abnormal respiration and estimate the Apnea-Hypopnea-Index (AHI) with a wearable respiratory device, compared to an SpO2 signal or polysomnography using a large (n = 412) dataset serving as ground truth.Methods: Simultaneously recorded polysomnographic (PSG) and wearable respiratory effort data were used to train and evaluate models in a cross-validation fashion. Time domain and complexity features were extracted, important features were identified, and a random forest model employed to detect events and predict AHI. Four models were trained: one each using the respiratory features only, a feature from the SpO2 (%)-signal only, and two additional models that use the respiratory features and the SpO2 (%)-feature, one allowing a time lag of 30 seconds between the two signals. Results:Event-based classification resulted in areas under the receiver operating characteristic curves of 0.94, 0.86, 0.82, and areas under the precision-recall curves of 0.48, 0.32, 0.51 for the models using respiration and SpO2, respiration-only, and SpO2-only respectively. Correlation between expert-labelled and predicted AHI was 0.96, 0.78, and 0.93, respectively. Conclusions:A wearable respiratory effort signal with or without SpO2 predicted AHI accurately. Given the large dataset and rigorous testing design, we expect our models are generalizable to evaluating respiration in a variety of environments, such as at home and in critical care.

The Role of Novel Digital Clinical Tools in the Screening or Diagnosis of Obstructive Sleep Apnea: Systematic Review

“…In Table 5 , we can see that, for the patient selection domain, only 1 study showed a high risk of bias [ 43 ], with 6 studies having a low risk [ 22 , 36 , 37 , 40 , 41 , 44 ], and 3 studies being unclear [ 38 , 39 , 42 ]. As for the index test domain, 6 studies presented a high risk [ 36 - 38 , 40 - 42 ] among questions Q5 and Q10, and the other 4 studies showed a low risk of bias [ 22 , 39 , 43 , 44 ].…”

“…The device used in 4 of them was a smartwatch, making use of their built-in sensors, such as photoplethysmography signals or an accelerometer [36][37][38][39]. For the other wearables, there was a garment or body-worn sensor that recorded electrocardiogram, thoracic movements, and positional signals [40]; a respiration belt that registered thoracic movement during respiration [41]; and a neck device that recorded snoring and position using a microphone and accelerometer, respectively [42]. In addition, 2 adhesive patches were included: one that recorded electrocardiogram and actigraphy signals that were patched to the chest [43], and one that recorded blood saturation on the hand using a photoelectric reflex sensor [22].…”

Association of Obstructive Sleep Apnea With Postoperative Delirium in Procedures of Moderate-to-High Complexity: A Hospital-Registry Study

“…Clinical applications and testing should be made available to patients to identify those at high risk of sleep apnea. 39 Future studies might target OSA patients as a high-risk population for delirium after complex procedures.…”

Smart Watches: Capturing Our Hearts

“…The gold standard of OSA diagnosis is through PSG, but a recent study has developed an algorithm to detect sleep apnea via smart watch utilizing motion data from the accelerometer with a precision of 0.97 and F1-score (accuracy) of 0.96 when compared to PSG [38]. As a defining feature of OSA is apneic periods, the addition of blood-oxygen measurement in the newest generation of Apple Watch presents a new, potentially useful parameter that may increase its accuracy in the detection of sleep apnea [39].…”