Apnea Detection in Polysomnographic Recordings Using Machine Learning Techniques

Sleep disorders are diagnosed in sleep laboratories by polysomnography, a multi-parameter examination that monitors biological signals during sleep. The subsequent evaluation of the obtained records is very time-consuming. The goal of this study was to create an automatic system for evaluation of the airflow and SpO2 channels of polysomnography records, through the use of machine learning techniques and a large database, for apnea and desaturation detection (which is unusual in other studies). To that end, a convolutional neural network (CNN) was designed using hyperparameter optimization. It was then trained and tested for apnea and desaturation. The proposed CNN was compared with the commonly used k-nearest neighbors (k-NN) method. The classifiers were designed based on nasal airflow and blood oxygen saturation signals. The final neural network accuracy for apnea detection reached 84%, and that for desaturation detection was 74%, while the k-NN classifier reached accuracies of 83% and 64% for apnea detection and desaturation detection, respectively.

P123 Accuracy and Reliability of the Transcutaneous Carbon Dioxide (TcCO2) Signal in the Sleep Laboratory

Carbon Dioxide (CO2) monitoring is an essential part of assessing and treating disorders of hypoventilation in the sleep laboratory. While reliablity issues have been previously reported with the Transcutaneous Carbon Dioxide (TcCO2) signal, there is limited data assessing the validity of this signal or its trend in the sleep laboratory context. Therefore, this study aimed to investigate the change in TcCO2 accuracy from the beginning to the end of the sleep study in real world conditions across two different Victorian public hospital sleep laboratories that used two different TcCO2 monitors. The sample included 13 consecutive patients from Monash Health and 44 consecutive patients from Alfred Health with an average age of 64 and 56 years respectively. Arterial Blood Gas (ABG) measurements were taken prior to and following each sleep study and compared concurrently with the TcCO2 value. Bland-Altman analysis revealed an average difference between TcCO2 and PaCO2 of 3.29mmHg with agreement between -11.44 and 16.64mmHg for the TCM4 device and 1.31mmHg with agreement between -7.64 and 9.05mmHg for the TCM5 device. When accuracy was compared across time points for each patient, 46% of patients had an overnight accuracy change of ≥ 8mmHg when using the TCM4 compared with 20% when using the TCM5. It was concluded that the TcCO2 signal was un-reliable across the different monitors and that the TcCO2 trend may be difficult to interpret with confidence without blood gas calibration at the commencement and conclusion of the sleep study.

P138 Nocturnal oxygen desaturation and MET call criteria during in-laboratory sleep studies

Introduction Patients undergoing sleep studies can experience frequent and profound oxygen desaturation. Most hospitals have standard MET (Medical Emergency Team) call criteria which obligate a response to severe oxygen desaturation. At our tertiary institution this is “Pulse oximetry/oxygen saturation: < 90 despite oxygen administration”. For most sleep studies provision of oxygen overnight would not be appropriate. We sought to examine the proportion of our sleep study patients who would meet MET call criteria. Method We retrospectively examined the data of all sleep studies which were performed in our laboratory between 01/01/2021 and 30/04/2021. Demographic and pulse oximetry data was collected. Results We collected data from 448 studies (95 CPAP, 342 diagnostic, 9 Split, 2 other). Patients were 40% female, 49±15 (mean±SD) years old and had a median AHI of 10 events per hour. 290 (65%) patients had a nadir SpO2 of <90%. The percentage of patients below with nadir SpO2 of 80%, 70%, 60% and 50% was 20%, 9%, 5% and 3% respectively. These proportions did not significantly change if treatment studies were excluded. In contrast, 23 (5%) of patients had a mean overnight SpO2 < 90%. During the period studied no serious adverse event was recorded. Discussion Most patients presenting for a sleep study to our tertiary institution would potentially meet standard hospital MET call criteria. This demonstrates the need for hospitals to be flexible in terms of hospital wide protocols when it comes to sleep laboratories. Evidence based criteria for medical escalation in sleep laboratories are required.

Assessment of effects of total sleep deprivation and subsequent recovery sleep: a methodological strategy feasible without sleep laboratory

Background Sleep is critical for maintaining homeostasis in bodily and neurobehavioral functions. This homeostasis can be disturbed by sleep interruption and restored to normal by subsequent recovery sleep. Most research regarding recovery sleep (RS) effects has been conducted in specialized sleep laboratories, whereas small, less-well equipped research units may lack the possibilities to run studies in this area. Hence, the aims of the present study were to develop and validate an experimental protocol, which allows a thorough assessment of at-home recovery sleep after sleep deprivation. Methods The experimental protocol, comprising one night of baseline sleep (BL) at home, one night of monitored total sleep deprivation and a subsequent recovery night at home, was tested in a sample of 30 healthy participants. Subjects’ fatigue and alertness were assessed prior to and after each night. Sleep at home (BL, RS) was objectively assessed using portable polysomnography. To check whether our at-home sleep assessments yielded results that are comparable to those conducted in sleep laboratories, we compared the sleep data assessed in our study with sleep data assessed in laboratory studies. Results Sleep parameters assessed during RS exhibited changes as expected (prolonged total sleep time, better sleep efficiency, slow wave sleep rebound). Sleep parameters of BL and RS were in line with parameters assessed in previous studies examining sleep in a laboratory setting. Fatigue normalized after one night of RS; alertness partly recovered. Conclusions Our results suggest a successful implementation of our new experimental protocol, emphasizing it as a useful tool for future studies on RS outside of well-equipped sleep laboratories.

Transforming electronic health record polysomnographic data into the Observational Medical Outcome Partnership's Common Data Model: a pilot feasibility study

Well-defined large-volume polysomnographic (PSG) data can identify subgroups and predict outcomes of obstructive sleep apnea (OSA). However, current PSG data are scattered across numerous sleep laboratories and have different formats in the electronic health record (EHR). Hence, this study aimed to convert EHR PSG into a standardized data format—the Observational Medical Outcome Partnership (OMOP) common data model (CDM). We extracted the PSG data of a university hospital for the period from 2004 to 2019. We designed and implemented an extract–transform–load (ETL) process to transform PSG data into the OMOP CDM format and verified the data quality through expert evaluation. We converted the data of 11,797 sleep studies into CDM and added 632,841 measurements and 9,535 observations to the existing CDM database. Among 86 PSG parameters, 20 were mapped to CDM standard vocabulary and 66 could not be mapped; thus, new custom standard concepts were created. We validated the conversion and usefulness of PSG data through patient-level prediction analyses for the CDM data. We believe that this study represents the first CDM conversion of PSG. In the future, CDM transformation will enable network research in sleep medicine and will contribute to presenting more relevant clinical evidence.

Sleep Laboratories Reopening and COVID-19: A European Perspective

The clinical activities regarding sleep disordered breathing (SDB) have been sharply interrupted during the initial phase of the COVID-19 epidemic throughout Europe. In the last months, activities have gradually restarted, according to epidemiological phase of COVID-19 and National recommendations. The recent increase in cases throughout Europe obliges to reconsider management strategies of SDB accordingly. Diagnosis of SDB and initiation of treatment pose some specific problems to be addressed to preserve safety of the patients and health personnel. This perspective document by a group of European sleep experts aims at summarising some different approaches followed in Europe and United States, which reflect National recommendations according to the epidemiological phase of the COVID-19 infection. Respiratory sleep medicine will likely change in the near future, and use of telemedicine will grow to avoid unnecessary risks and continue to provide optimal care to the patients. The document also covers pediatric sleep studies and indications for titration of noninvasive ventilation, as well as precautions to be followed by patients who are already on positive airway pressure treatment. A single consensus document developed by the European Respiratory Society and National Societies would be desirable to harmonise SDB management throughout Europe.

New organisation for follow-up and assessment of treatment efficacy in sleep apnoea

Obstructive sleep apnoea (OSA) is a highly prevalent disease, and there is an increased demand for OSA diagnosis and treatment. However, resources are limited compared with the growing needs for OSA diagnosis and management, and alternative strategies need to be developed to optimise the OSA clinical pathway. In this review, we propose a management strategy for OSA, and in general for sleep-disordered breathing, to be implemented from diagnosis to follow-up. For this purpose, the best current options seem to be: 1) networking at different levels of care, from primary physicians to specialised sleep laboratories; and 2) use of telemedicine. Telemedicine can contribute to the improved cost-effectiveness of OSA management during both the diagnostic and therapeutic phases. However, although the technology is already in place and different commercial platforms are in use, it is still unclear how to use telemedicine effectively in the sleep field. Application of telemedicine for titration of positive airway pressure treatment, follow-up to improve compliance to treatment through early identification and solution of problems, and teleconsultation all appear to be promising areas for improved OSA management.

Cloud algorithm-driven oximetry-based diagnosis of obstructive sleep apnoea in symptomatic habitually snoring children

The ability of a cloud-driven Bluetooth oximetry-based algorithm to diagnose obstructive sleep apnoea syndrome (OSAS) was examined in habitually snoring children concurrently undergoing overnight polysomnography.Children clinically referred for overnight in-laboratory polysomnographic evaluation for suspected OSAS were simultaneously hooked to a Bluetooth oximeter linked to a smartphone. Polysomnography findings were scored and the apnoea/hypopnoea index (AHIPSG) was tabulated, while oximetry data yielded an estimated AHIOXI using a validated algorithm.The accuracy of the oximeter in identifying correctly patients with OSAS in general, or with mild (AHI 1–5 events·h−1), moderate (5–10 events·h−1) or severe (>10 events·h−1) OSAS was examined in 432 subjects (6.5±3.2 years), with 343 having AHIPSG >1 event·h−1. The accuracies of AHIOXI were consistently >79% for all levels of OSAS severity, and specificity was particularly favourable for AHI >10 events·h−1 (92.7%). Using the criterion of AHIPSG >1 event·h−1, only 4.7% of false-negative cases emerged, from which only 0.6% of cases showed moderate or severe OSAS.Overnight oximetry processed via Bluetooth technology by a cloud-based machine learning-derived algorithm can reliably diagnose OSAS in children with clinical symptoms suggestive of the disease. This approach provides virtually limitless scalability and should alleviate the substantial difficulties in accessing paediatric sleep laboratories while markedly reducing the costs of OSAS diagnosis.

Setting Up a Pediatric Sleep Lab

Obstructive sleep apnea occurs in a significant proportion of children and adolescents and requires a sleep study to diagnose the condition. However, there are relatively few sleep laboratories that serve this population. Consequently, this means sleep studies are not done in a timely manner, and many of these patients do not get studies performed when indicated. Building new pediatric-focused sleep laboratories or expanding service in an adult-focused laboratory to children can help overcome this barrier.The decision to build or modify an existing sleep laboratory for children brings many considerations that are different than for adults. The location of the laboratory is partially determined by the need for the presence of a sleep technologist. Whether they are done in the community or a hospital will be affected by the patient's medical complexity. The design of the sleep laboratory can also be influenced by the presence of children. All children, under 18 years of age, will require a parent to sleep in the room with them. Safety will also be impacted. For example, electric outlets need to be protected, furniture should be child safe, and transportation to emergency facilities must be managed. In addition, service to children also raises technical issues. They require different types of leads and smaller equipment and the software must meet required pediatric specifications. The staff must understand pediatric developmental, social, and medical needs. It is also critical that they have a desire to work with children.This article is written to assist the reader in building a sleep laboratory with the pediatric patient in mind.