Predicting Obstructive Sleep Apnea Based on Computed Tomography Scan Using Deep Learning Models.

RATIONALE: The incidence of clinically undiagnosed obstructive sleep apnea (OSA) is high among the general population due to limited access to polysomnography. Computed tomography (CT) of craniofacial regions obtained for other purposes can be beneficial in predicting OSA and its severity. OBJECTIVES: To predict OSA and its severity based on paranasal CT using a 3-dimensional deep learning algorithm. METHODS: One internal dataset (n=798) and two external datasets (n=135 and 85) were used in this study. In the internal dataset, 92 normal, 159 mild, 201 moderate, and 346 severe OSA participants were enrolled to derive the deep learning model. A multimodal deep learning model was elicited from the connection between a 3-dimensional convolutional neural network (CNN)-based part treating unstructured data (CT images) and a multi-layer perceptron (MLP)-based part treating structured data (age, sex, and body mass index) to predict OSA and its severity. MEASUREMENTS AND MAIN RESULTS: In four-class classification for predicting the severity of OSA, the AirwayNet-MM-H model (multimodal model with airway-highlighting preprocessing algorithm) showed an average accuracy of 87.6% (95% confidence interval [CI] 86.8-88.6) in the internal dataset and 84.0% (95% CI 83.0-85.1) and 86.3% (95% CI 85.3-87.3) in the two external datasets, respectively. In the two-class classification for predicting significant OSA (moderate to severe OSA), The area under the receiver operating characteristics (AUROC), accuracy, sensitivity, specificity, and F1 score were 0.910 (95% CI 0.899-0.922), 91.0% (95% CI 90.1-91.9), 89.9% (95% CI 88.8-90.9), 93.5% (95% CI 92.7-94.3), and 93.2% (95% CI 92.5-93.9), respectively, in the internal dataset. Furthermore, the diagnostic performance of the Airway Net-MM-H model outperformed that of the other six state-of-the-art deep learning models in terms of accuracy for both four- and two-class classifications and AUROC for two-class classification (p<0.001). CONCLUSIONS: A novel deep learning model, including a multimodal deep learning model and an airway-highlighting preprocessing algorithm from CT images obtained for other purposes, can provide significantly precise outcomes for OSA diagnosis.

How to objectively measure snoring: a systematic review.

PURPOSE: Snoring is the most common symptom of obstructive sleep apnea. Various objective methods of measuring snoring are available, and even if the measurement is performed the same way, communication is difficult because there are no common reference values between the researcher and clinician with regard to intensity and frequency, among other variables. In other words, no consensus regarding objective measurement has been reached. This study aimed to review the literature related to the objective measurement of snoring, such as measurement devices, definitions, and device locations. METHODS: A literature search based on the PubMed, Cochrane, and Embase databases was conducted from the date of inception to April 5, 2023. Twenty-nine articles were included in this study. Articles that mentioned only the equipment used for measurement and did not include individual details were excluded from the study. RESULTS: Three representative methods for measuring snoring emerged. These include (1) a microphone, which measures snoring sound; (2) piezoelectric sensor, which measures snoring vibration; and (3) nasal transducer, which measures airflow. In addition, recent attempts have been made to measure snoring using smartphones and applications. CONCLUSION: Numerous studies have investigated both obstructive sleep apnea and snoring. However, the objective methods of measuring snoring and snoring-related concepts vary across studies. Consensus in the academic and clinical communities on how to measure and define snoring is required.

Real-Time Detection of Sleep Apnea Based on Breathing Sounds and Prediction Reinforcement Using Home Noises: Algorithm Development and Validation.

BACKGROUND: Multinight monitoring can be helpful for the diagnosis and management of obstructive sleep apnea (OSA). For this purpose, it is necessary to be able to detect OSA in real time in a noisy home environment. Sound-based OSA assessment holds great potential since it can be integrated with smartphones to provide full noncontact monitoring of OSA at home. OBJECTIVE: The purpose of this study is to develop a predictive model that can detect OSA in real time, even in a home environment where various noises exist. METHODS: This study included 1018 polysomnography (PSG) audio data sets, 297 smartphone audio data sets synced with PSG, and a home noise data set containing 22,500 noises to train the model to predict breathing events, such as apneas and hypopneas, based on breathing sounds that occur during sleep. The whole breathing sound of each night was divided into 30-second epochs and labeled as "apnea," "hypopnea," or "no-event," and the home noises were used to make the model robust to a noisy home environment. The performance of the prediction model was assessed using epoch-by-epoch prediction accuracy and OSA severity classification based on the apnea-hypopnea index (AHI). RESULTS: Epoch-by-epoch OSA event detection showed an accuracy of 86% and a macro F1-score of 0.75 for the 3-class OSA event detection task. The model had an accuracy of 92% for "no-event," 84% for "apnea," and 51% for "hypopnea." Most misclassifications were made for "hypopnea," with 15% and 34% of "hypopnea" being wrongly predicted as "apnea" and "no-event," respectively. The sensitivity and specificity of the OSA severity classification (AHI≥15) were 0.85 and 0.84, respectively. CONCLUSIONS: Our study presents a real-time epoch-by-epoch OSA detector that works in a variety of noisy home environments. Based on this, additional research is needed to verify the usefulness of various multinight monitoring and real-time diagnostic technologies in the home environment.

Prediction of Sleep Stages Via Deep Learning Using Smartphone Audio Recordings in Home Environments: Model Development and Validation.

BACKGROUND: The growing public interest and awareness regarding the significance of sleep is driving the demand for sleep monitoring at home. In addition to various commercially available wearable and nearable devices, sound-based sleep staging via deep learning is emerging as a decent alternative for their convenience and potential accuracy. However, sound-based sleep staging has only been studied using in-laboratory sound data. In real-world sleep environments (homes), there is abundant background noise, in contrast to quiet, controlled environments such as laboratories. The use of sound-based sleep staging at homes has not been investigated while it is essential for practical use on a daily basis. Challenges are the lack of and the expected huge expense of acquiring a sufficient size of home data annotated with sleep stages to train a large-scale neural network. OBJECTIVE: This study aims to develop and validate a deep learning method to perform sound-based sleep staging using audio recordings achieved from various uncontrolled home environments. METHODS: To overcome the limitation of lacking home data with known sleep stages, we adopted advanced training techniques and combined home data with hospital data. The training of the model consisted of 3 components: (1) the original supervised learning using 812 pairs of hospital polysomnography (PSG) and audio recordings, and the 2 newly adopted components; (2) transfer learning from hospital to home sounds by adding 829 smartphone audio recordings at home; and (3) consistency training using augmented hospital sound data. Augmented data were created by adding 8255 home noise data to hospital audio recordings. Besides, an independent test set was built by collecting 45 pairs of overnight PSG and smartphone audio recording at homes to examine the performance of the trained model. RESULTS: The accuracy of the model was 76.2% (63.4% for wake, 64.9% for rapid-eye movement [REM], and 83.6% for non-REM) for our test set. The macro F1-score and mean per-class sensitivity were 0.714 and 0.706, respectively. The performance was robust across demographic groups such as age, gender, BMI, or sleep apnea severity (accuracy 73.4%-79.4%). In the ablation study, we evaluated the contribution of each component. While the supervised learning alone achieved accuracy of 69.2% on home sound data, adding consistency training to the supervised learning helped increase the accuracy to a larger degree (+4.3%) than adding transfer learning (+0.1%). The best performance was shown when both transfer learning and consistency training were adopted (+7.0%). CONCLUSIONS: This study shows that sound-based sleep staging is feasible for home use. By adopting 2 advanced techniques (transfer learning and consistency training) the deep learning model robustly predicts sleep stages using sounds recorded at various uncontrolled home environments, without using any special equipment but smartphones only.

A Classifying Model of Obstructive Sleep Apnea Based on Heart Rate Variability in a Large Korean Population.

BACKGROUND: The majority of patients with obstructive sleep apnea do not receive timely diagnosis and treatment because of the complexity of a diagnostic test. We aimed to predict obstructive sleep apnea based on heart rate variability, body mass index, and demographic characteristics in a large Korean population. METHODS: Models of binary classification for predicting obstructive sleep apnea severity were constructed using 14 features including 11 heart rate variability variables, age, sex, and body mass index. Binary classification was conducted separately using apnea-hypopnea index thresholds of 5, 15, and 30. Sixty percent of the participants were randomly allocated to training and validation sets while the other forty percent were designated as the test set. Classifying models were developed and validated with 10-fold cross-validation using logistic regression, random forest, support vector machine, and multilayer perceptron algorithms. RESULTS: A total of 792 (651 men and 141 women) subjects were included. The mean age, body mass index, and apnea-hypopnea index score were 55.1 years, 25.9 kg/m², and 22.9, respectively. The sensitivity of the best performing algorithm was 73.6%, 70.7%, and 78.4% when the apnea-hypopnea index threshold criterion was 5, 10, and 15, respectively. The prediction performances of the best classifiers at apnea-hypopnea indices of 5, 15, and 30 were as follows: accuracy, 72.2%, 70.0%, and 70.3%; specificity, 64.6%, 69.2%, and 67.9%; area under the receiver operating characteristic curve, 77.2%, 73.5%, and 80.1%, respectively. Overall, the logistic regression model using the apnea-hypopnea index criterion of 30 showed the best classifying performance among all models. CONCLUSION: Obstructive sleep apnea was fairly predicted by using heart rate variability, body mass index, and demographic characteristics in a large Korean population. Prescreening and continuous treatment monitoring of obstructive sleep apnea may be possible simply by measuring heart rate variability.

The 24-Hour Cardiac Autonomic Activity in Patients With Allergic Rhinitis.

BACKGROUND: Definitive knowledge of the 24-hour cardiac autonomic activity in patients with allergic rhinitis (AR) is lacking. Thus, we aimed to evaluate heart rate variability (HRV), which is used to measure cardiac autonomic activity by 24-hour Holter monitoring in patients with AR. METHODS: We enrolled 32 patients who visited our clinic and were diagnosed with AR. The control group was selected four-fold (n = 128) by matching (age, sex, hypertension, and diabetes) in the AR group from a Holter registry in the cardiology department. The HRV results, which were measured using 24-hour Holter monitoring, were compared between the AR and control groups. RESULTS: All time-domain parameters of HRV revealed no differences between the groups. However, among the frequency domain parameters of HRV, the low-frequency to high-frequency ratio and low-frequency power in normalized units were significantly lower in the AR group. Conversely, high-frequency power in normalized units was significantly higher in the AR group. In the multiple regression analysis, AR was independently associated with sympathetic withdrawal (adjusted odds ratio = 3.393, P = 0.020) after adjusting for age, sex, hypertension, diabetes mellitus, and hyperlipidemia. CONCLUSIONS: The present findings suggest differences in cardiac autonomic activity which are related with sympathetic withdrawal in patients with AR compared with that in the normal population over 24 hours.

Correlation of site of obstruction between two dynamic evaluation modalities in obstructive sleep apnea patients: drug-induced sleep endoscopy and sleep videofluoroscopy.

PURPOSE: Drug-induced sleep endoscopy (DISE) and sleep videofluoroscopy (SVF) are two dynamic modalities for evaluating the upper airway in patients with obstructive sleep apnea (OSA). We evaluated the correlation of obstructive sites determined by DISE and SVF in OSA patients and elucidate findings that can improve the accuracy of upper airway assessment. METHODS: A consecutive series of 63 patients with OSA who underwent DISE and SVF were the subjects of this study. The DISE and SVF findings were divided according to the anatomical structure responsible for the collapse, including the soft palate (SP), oropharyngeal lateral walls (LW), tongue base (TB), and larynx (LX). The obstruction was graded on the three-point scale: 0, no obstruction; 1, partial obstruction; or 2, complete obstruction. Additionally, grade 1.5 TB obstruction was designated when the posterior displacement of the anterior tongue was detected during simultaneous retropalatal obstruction. The agreement rate and Cohen's kappa test between the two modalities were also assessed. RESULTS: The agreement rate between the two modalities was highest in LX (88.9%) followed by SP (85.7%), TB (76.1%), and LW (74.6%) (Cohen's kappa value = 0.757 in LX, 0.642 in SP, 0.637 in TB, 0.612 in LW, respectively). When grade 1.5 and 2 TB obstructions were combined, the agreement rate increased to 88.9% (Cohen's kappa value = 0.757). CONCLUSIONS: We found a good overall agreement between the two dynamic airway evaluation modalities during drug-induced sleep, and this correlation may be improved if the posterior displacement of the anterior tongue during DISE is used as a sign of TB obstruction.

Regional and Chronological Variation of Chemosensory Dysfunction in COVID-19: a Meta-Analysis.

BACKGROUND: Olfactory and gustatory dysfunction are frequently reported in patients with coronavirus disease 2019 (COVID-19). However, the reported prevalence of olfactory and/or gustatory dysfunction varies widely, and the reason for the inter-study differences is unclear. Hence, in this meta-analysis, we performed subgroup analyses to investigate the factors that contribute to the inter-study variability in the prevalence of olfactory and gustatory dysfunction. METHODS: Out of 943 citations, we included 55 eligible studies with 13,527 patients with COVID-19 for a meta-analysis. Calculating the data extracted from each study, the weighted summary prevalence of olfactory and gustatory dysfunction was estimated using a Freeman-Tukey transformation with models based on random-effects assumptions. A meta-analysis of variance compared the prevalence of olfactory and gustatory dysfunction according to regional, chronological, demographic, and methodologic factors, respectively. RESULTS: The overall pooled prevalence rates of olfactory and gustatory dysfunction were 51.4% and 47.5%, respectively, in the random-effect model. In subgroup analyses, the prevalence rates of olfactory and gustatory dysfunction were significantly different among four geographical regions (both P < 0.001, respectively). Although the prevalence rates of olfactory and gustatory dysfunction did not significantly differ according to the time of enrollment, the subgroup analyses including only studies from the same geographical region (Europe) revealed a significant difference in olfactory dysfunction according to the time of enrollment. CONCLUSION: The regional and chronological differences in the prevalence rates of olfactory and gustatory dysfunctions partly explain the wide inter-study variability.

Validation of Olfactory Questionnaire in Koreans: an Alternative for Conventional Psychophysical Olfactory Tests.

BACKGROUND: The patients with coronavirus disease 2019 (COVID-19), a worldwide pandemic infection, frequently complain of olfactory disorders. However, psychophysical olfactory tests performed by an examiner are very difficult in these highly infectious patients. This study aimed to develop and validate a questionnaire for olfactory function that can be readily used to evaluate olfactory loss. METHODS: Fourteen smell-related questions were created based on smells familiar to Koreans. Among them, questions with a κ value of 0.6 or higher were finally selected through a test-retest reliability analysis. The correlations between the scores of the olfactory questionnaire and those of olfactory function tests (Butanol Threshold Test [BTT] and Cross Cultural Smell Identification Test [CCSIT]) were analyzed. To evaluate the predictive ability of the questionnaire and elicit cutoff values, receiver operating characteristic (ROC) curves were generated. RESULTS: Out of the 14 questions in the questionnaire, 11 (κ > 0.6) were selected for the olfactory questionnaire. We analyzed 2,273 subjects, and there was a significant correlation between the total score of the olfactory questionnaire and the BTT (r = 0.643, P < 0.001) or CCSIT (r = 0.615, P < 0.001) scores. ROC curves for the olfactory questionnaire, BTT, and CCSIT all demonstrated high predictive power to discriminate anosmia and severe hyposmia from normosmia. Regarding mild to moderate hyposmia, however, ROC curve for the olfactory questionnaire alone showed high predictive power of discrimination from normosmia. Based on the results of ROC curves among the subclasses, we suggest the classification of the total score of the questionnaire as 0-4, 5-17, 18-27, 28-41, and 42-44, for anosmia, severe hyposmia, moderate hyposmia, mild hyposmia, and normosmia, respectively. CONCLUSION: The total scores of the questionnaires correlated with the BTT and CCSIT scores. The symptom questionnaire for olfactory dysfunction may be useful as an alternative tool for olfactory function testing, when unavailable.

Risk Model Establishment of Endoscopic Sinus Surgery for Patients with Chronic Rhinosinusitis: a Multicenter Study in Korea.

BACKGROUND: Endoscopic sinus surgery (ESS) is the mainstay treatment for refractory chronic rhinosinusitis (CRS). Since various factors may contribute to the surgical outcome, it is challenging for physicians to predict surgical outcomes. The aim of study was to analyze the prognostic factors of postoperative outcomes and to establish the prediction model with the risk factors that impact the postoperative outcomes. METHODS: Medical records of CRS patients who underwent ESS at 9 institutions in 2005, 2010, and 2016 were retrospectively reviewed. We classified the patients into 2 groups based on postoperative objective endoscopic outcomes. Demographics, nose-specific symptoms, olfactory function, eosinophil counts in blood (EoB) and nasal tissue (EoT), and Lund-Mackay CT score (LMS) were collected. Univariate and multivariate analyses were performed and established a prediction equation for postoperative endoscopic objective outcomes. RESULTS: In total (n = 1,249), 27.0% were not satisfied under postoperative endoscopic examination. Of 10 variables, LMS (> 5), sinus dominancy (maxillary sinus and ethmoid sinus), EoB (> 210), and EoT (> 100) were statistically significant in univariate analysis (P < 0.05, all). In multivariate analysis, EoT (> 100) and LMS (> 5) were significantly associated with poor postoperative outcome. Furthermore, 5 significant variables were employed to establish the risk model of postoperative outcomes and P (the value of prediction probability) = 1 / (1 + exp [-0.392 + 1.088 × EoT (> 100) + 0.123 × mean LMS (> 5) - 0.366 × sinus dominancy (maxillary) + 0.064 × sinus dominancy (similar) + 0.200 × EoB (4%) + 0.344 × EoB (> 210)] was developed. CONCLUSION: Tissue eosinophil count and radiographic severity predispose to a poorer outcome of ESS and the risk model established may be helpful to predict postoperative outcomes of ESS.

Efficacy and safety of rupatadine fumarate in the treatment of perennial allergic rhinitis: A multicenter, double-blinded, randomized, placebo-controlled, bridging study in Koreans.

BACKGROUND: The efficacy of rupatadine for the treatment of AR has been confirmed in numerous clinical studies, however there are very few studies on asian patients. OBJECTIVE: To assess the safety and efficacy of rupatadine fumarate in the treatment of Korean perennial allergic rhinitis (PAR) patients. METHODS: A multicenter, double-blind, randomized, placebo-controlled, comparative study of rupatadine fumarate and bepotastine besilate was conducted. Each group was administered rupatadine, bepotastine or placebo for 4 weeks. Primary parameters for efficacy included morning and evening symptom reduction from baseline at 4 weeks. Treatment safety and tolerability were evaluated according to a self-reported incidence and type of adverse events at each follow up visit. RESULTS: Rupatadine showed a significant reduction in symptoms at morning and evening evaluations, in both 5TSS (-5.69, P < 0.0006) and 4NTSS (-4.74, P < 0.0015) compared to placebo. There was a significant reduction from baseline for 5TSS (-65.4%, P = 0.002) and 4NTSS (-63.7%, P = 0.003) with rupatadine compared with placebo. At evening evaluations, there were significant reductions of 5TSS (-63.2%, P = 0.009) and 4NTSS (-61.6%, P = 0.013) for the rupatadine group. Compared with bepotastine, rupatadine showed greater reduction in the morning symptoms at 4 weeks. When individual symptoms were assessed with 12-hour reflective mean daily symptom score, rupatadine showed better efficacy than placebo in sneezing (P = 0.016) and rhinorrhea (P = 0.097). The rate of adverse events showed no statistical significance. CONCLUSIONS: Rupatadine is a safe and effective treatment option for Korean PAR patients and possibly a better choice over bepotastine for controlling morning symptom.

Converting clinical document architecture documents to the common data model for incorporating health information exchange data in observational health studies: CDA to CDM.

BACKGROUND: Utilization of standard health information exchange (HIE) data is growing due to the high adoption rate and interoperability of electronic health record (EHR) systems. However, integration of HIE data into an EHR system is not yet fully adopted in clinical research. In addition, data quality should be verified for the secondary use of these data. Thus, the aims of this study were to convert referral documents in a Health Level 7 (HL7) clinical document architecture (CDA) to the common data model (CDM) to facilitate HIE data availability for longitudinal data analysis, and to identify data quality levels for application in future clinical studies. METHODS: A total of 21,492 referral CDA documents accumulated for over 10 years in a tertiary general hospital in South Korea were analyzed. To convert CDA documents to the Observational Medical Outcomes Partnership (OMOP) CDM, processes such as CDA parsing, data cleaning, standard vocabulary mapping, CDA-to-CDM mapping, and CDM conversion were performed. The quality of CDM data was then evaluated using the Achilles Heel and visualized with the Achilles tool. RESULTS: Mapping five CDA elements (document header, problem, medication, laboratory, and procedure) into an OMOP CDM table resulted in population of 9 CDM tables (person, visit_occurrence, condition_occurrence, drug_exposure, measurement, observation, procedure_occurrence, care_site, and provider). Three CDM tables (drug_era, condition_era, and observation_period) were derived from the converted table. From vocabulary mapping codes in CDA documents according to domain, 98.6% of conditions, 68.8% of drugs, 35.7% of measurements, 100% of observation, and 56.4% of procedures were mapped as standard concepts. The conversion rates of the CDA to the OMOP CDM were 96.3% for conditions, 97.2% for drug exposure, 98.1% for procedure occurrence, 55.1% for measurements, and 100% for observation. CONCLUSIONS: We examined the possibility of CDM conversion by defining mapping rules for CDA-to-CDM conversion using the referral CDA documents collected from clinics in actual medical practice. Although mapping standard vocabulary for CDM conversion requires further improvement, the conversion could facilitate further research on the usage patterns of medical resources and referral patterns.

Effectiveness of sleep surgery versus a mandibular advancement device for obstructive sleep apnea in terms of nocturnal cardiac autonomic activity.

PURPOSE: Sleep surgery and mandibular advancement devices (MAD) are treatments for obstructive sleep apnea (OSA), but their comparative efficacy remains unclear. We compared their efficacy using various parameters. METHODS: Subjects treated for OSA with sleep surgery or MAD (n = 30/group)-matched for sex, body mass index (BMI), and baseline apnea-hypopnea index (AHI)-were enrolled. The efficacy of these treatments according to polysomnographic parameters, sleep quality questionnaires, and heart rate variability (HRV) time- and frequency-domain parameters were compared between pre-treatment and 3-month post-treatment. RESULTS: Polysomnographic and sleep quality questionnaire parameters improved significantly in both groups. In time-domain HRV analysis, average normal-to-normal intervals increased significantly in the surgery (942.2 ± 140.8 to 994.6 ± 143.1, P = 0.008) and MAD (901.1 ± 131.7 to 953.7 ± 123.1, P = 0.002) groups. Low frequency (LF) decreased significantly in the surgery group (P = 0.012); high frequency (HF) remained unchanged in both groups. The LF/HF ratio decreased in both groups (2.9 ± 1.8 to 2.3 ± 1.7, P = 0.017, vs. 3.0 ± 1.8 to 2.4 ± 1.4, P = 0.025). Normalized high frequency increased significantly in both groups (31.0 ± 13.2 to 36.8 ± 13.7, P = 0.009, vs. 29.1 ± 10.7 to 33.7 ± 12.5, P = 0.024), in contrast to normalized low frequency. However, no HRV parameter changes differed significantly between the groups after adjusting for age, BMI, and AHI. CONCLUSION: Sleep surgery and MAD are equally effective treatments for OSA according to cardiac autonomic activity.

Clinical Genomic Sequencing Reports in Electronic Health Record Systems Based on International Standards: Implementation Study.

BACKGROUND: To implement standardized machine-processable clinical sequencing reports in an electronic health record (EHR) system, the International Organization for Standardization Technical Specification (ISO/TS) 20428 international standard was proposed for a structured template. However, there are no standard implementation guidelines for data items from the proposed standard at the clinical site and no guidelines or references for implementing gene sequencing data results for clinical use. This is a significant challenge for implementation and application of these standards at individual sites. OBJECTIVE: This study examines the field utilization of genetic test reports by designing the Health Level 7 (HL7) Fast Healthcare Interoperability Resources (FHIR) for genomic data elements based on the ISO/TS 20428 standard published as the standard for genomic test reports. The goal of this pilot is to facilitate the reporting and viewing of genomic data for clinical applications. FHIR Genomics resources predominantly focus on transmitting or representing sequencing data, which is of less clinical value. METHODS: In this study, we describe the practical implementation of ISO/TS 20428 using HL7 FHIR Genomics implementation guidance to efficiently deliver the required genomic sequencing results to clinicians through an EHR system. RESULTS: We successfully administered a structured genomic sequencing report in a tertiary hospital in Korea based on international standards. In total, 90 FHIR resources were used. Among 41 resources for the required fields, 26 were reused and 15 were extended. For the optional fields, 28 were reused and 21 were extended. CONCLUSIONS: To share and apply genomic sequencing data in both clinical practice and translational research, it is essential to identify the applicability of the standard-based information system in a practical setting. This prototyping work shows that reporting data from clinical genomics sequencing can be effectively implemented into an EHR system using the existing ISO/TS 20428 standard and FHIR resources.

Clinicopathologic characteristics of paranasal sinus fungus ball: retrospective, multicenter study in Korea.

PURPOSE: Fungus ball (FB) is the most common type of fungal rhinosinusitis and the prevalence of FB has increased over the past 10 years. The aim of this study was to compare the clinical characteristics of Korean adult patients with FB and chronic rhinosinusitis (CRS) without FB. METHODS: We retrospectively analyzed data on 1362 patients (147 FB and 1215 CRS) who underwent endoscopic sinus surgery at nine Korean medical centers in 2005, 2010, and 2016. We evaluated the prevalence of FB and compared the clinical characteristics of FB and CRS. Medical records, computed tomography (CT) findings, atopic status, concomitant diseases, tissue, and blood eosinophil count were assessed. RESULTS: The prevalence of FB was significantly higher in 2016 (15.9%) than in the other years (7.8% in 2005 and 7.5% in 2010). The FB patients were more likely to be female, older, have unilateral disease and less likely to have allergy compared to the CRS patients. The most common main complaint related to CRS and FB was nasal obstruction. CT determined that unilateral disease and maxillary sinus dominancy were common in patients with FB. The incidence of concomitant diseases was much higher in FB, with lower tissue and blood eosinophilia. CONCLUSION: FB is commonly encountered in older women with the increased prevalence. FB had a different clinical presentation, radiological findings, and prognosis than CRS. Further studies are needed to understand the pathophysiologic mechanisms underlying the development of FB.

Patient-Level Prediction of Cardio-Cerebrovascular Events in Hypertension Using Nationwide Claims Data.

BACKGROUND: Prevention and management of chronic diseases are the main goals of national health maintenance programs. Previously widely used screening tools, such as Health Risk Appraisal, are restricted in their achievement this goal due to their limitations, such as static characteristics, accessibility, and generalizability. Hypertension is one of the most important chronic diseases requiring management via the nationwide health maintenance program, and health care providers should inform patients about their risks of a complication caused by hypertension. OBJECTIVE: Our goal was to develop and compare machine learning models predicting high-risk vascular diseases for hypertensive patients so that they can manage their blood pressure based on their risk level. METHODS: We used a 12-year longitudinal dataset of the nationwide sample cohort, which contains the data of 514,866 patients and allows tracking of patients' medical history across all health care providers in Korea (N=51,920). To ensure the generalizability of our models, we conducted an external validation using another national sample cohort dataset, comprising one million different patients, published by the National Health Insurance Service. From each dataset, we obtained the data of 74,535 and 59,738 patients with essential hypertension and developed machine learning models for predicting cardiovascular and cerebrovascular events. Six machine learning models were developed and compared for evaluating performances based on validation metrics. RESULTS: Machine learning algorithms enabled us to detect high-risk patients based on their medical history. The long short-term memory-based algorithm outperformed in the within test (F1-score=.772, external test F1-score=.613), and the random forest-based algorithm of risk prediction showed better performance over other machine learning algorithms concerning generalization (within test F1-score=.757, external test F1-score=.705). Concerning the number of features, in the within test, the long short-term memory-based algorithms outperformed regardless of the number of features. However, in the external test, the random forest-based algorithm was the best, irrespective of the number of features it encountered. CONCLUSIONS: We developed and compared machine learning models predicting high-risk vascular diseases in hypertensive patients so that they may manage their blood pressure based on their risk level. By relying on the prediction model, a government can predict high-risk patients at the nationwide level and establish health care policies in advance.

Detection of sleep disordered breathing severity using acoustic biomarker and machine learning techniques.

PURPOSE: Breathing sounds during sleep are altered and characterized by various acoustic specificities in patients with sleep disordered breathing (SDB). This study aimed to identify acoustic biomarkers indicative of the severity of SDB by analyzing the breathing sounds collected from a large number of subjects during entire overnight sleep. METHODS: The participants were patients who presented at a sleep center with snoring or cessation of breathing during sleep. They were subjected to full-night polysomnography (PSG) during which the breathing sound was recorded using a microphone. Then, audio features were extracted and a group of features differing significantly between different SDB severity groups was selected as a potential acoustic biomarker. To assess the validity of the acoustic biomarker, classification tasks were performed using several machine learning techniques. Based on the apnea-hypopnea index of the subjects, four-group classification and binary classification were performed. RESULTS: Using tenfold cross validation, we achieved an accuracy of 88.3% in the four-group classification and an accuracy of 92.5% in the binary classification. Experimental evaluation demonstrated that the models trained on the proposed acoustic biomarkers can be used to estimate the severity of SDB. CONCLUSIONS: Acoustic biomarkers may be useful to accurately predict the severity of SDB based on the patient's breathing sounds during sleep, without conducting attended full-night PSG. This study implies that any device with a microphone, such as a smartphone, could be potentially utilized outside specialized facilities as a screening tool for detecting SDB.

Evaluation of the success of obstructive sleep apnea surgery using criteria based on long-term symptoms and incident hypertension.

OBJECTIVE: To identify appropriate success criteria, based on long-term symptoms and incident hypertension, after surgery for obstructive sleep apnea (OSA). METHODS: This observational cohort study included 97 adult OSA patients (90 men) who underwent surgical treatment at our tertiary medical center. Subjective symptoms [witnessed sleep apnea and snoring, and Epworth sleepiness scale (ESS) scores] were evaluated through a telephone survey, and incident hypertension was assessed from medical records. The subjects were divided into success and failure groups according to seven different criteria, and data were analyzed to identify the criteria that could significantly differentiate the success from failure groups. RESULTS: The participants had a mean age of 48.8 ± 11.9 years and a mean preoperative body mass index of 26.5 ± 3.5 kg/m2. The mean preoperative and postoperative apnea-hypopnea index (AHI) values were 36.1/h and 19.4/h, respectively. The mean follow-up duration was 77.0 ± 31.1 months. Postoperative witnessed apnea, snoring, and the ESS scores decreased significantly compared to preoperative scores in both the success and failure groups based on most of the seven criteria. Among the seven criteria, success and failure groups based on a postoperative AHI cutoff of 15 or 20/h differed significantly in witnessed apnea, snoring, or ESS scores. Kaplan-Meier survival analysis based on incident hypertension revealed that no criterion could significantly distinguish between the two groups. CONCLUSIONS: Our results suggest that some of the success criteria analyzed may be more useful in differentiating between success and failure groups after surgery, in terms of long-term improvement of subjective OSA-related symptoms.

Soft palate cephalometric changes with a mandibular advancement device may be associated with polysomnographic improvement in obstructive sleep apnea.

PURPOSE: It is unclear whether soft palate-associated changes in cephalometry associated with a mandibular advancement device (MAD) are independently associated with improvements in polysomnography (PSG) respiratory parameters in obstructive sleep apnea (OSA). METHODS: This retrospective review aimed to identify the association between soft palate-associated changes in cephalometry and PSG changes after application of an MAD. Korean patients diagnosed with OSA who underwent cephalometry with or without an MAD were enrolled. All the patients were evaluated after undergoing full-night PSG twice: once with an MAD and once without. Cephalometric findings were measured using an image analyzer. RESULTS: Mean apnea-hypopnea index significantly decreased with an MAD from 36.4/h to 14.7/h (p < 0.001). Retropalatal airway space significantly increased with an MAD from 6.6 to 7.3 mm (p = 0.013). Soft palate length also significantly decreased with an MAD from 43.6 to 42.3 mm (p = 0.02). Although these findings were shown by responders (patients with a reduction of apnea-hypopnea index by more than 50%), there were no significant changes in non-responders. However, retroglossal airway space did not significantly increase with an MAD even in responders. CONCLUSIONS: Improvement of OSA with an MAD can be predicted with soft palate-associated upper airway changes shown in cephalometry.

Exploiting temporal and nonstationary features in breathing sound analysis for multiple obstructive sleep apnea severity classification.

BACKGROUND: Polysomnography (PSG) is the gold standard test for obstructive sleep apnea (OSA), but it incurs high costs, requires inconvenient measurements, and is limited by a one-night test. Thus, a repetitive OSA screening test using affordable data would be effective both for patients interested in their own OSA risk and in-hospital PSG. The purpose of this research was to develop a four-OSA severity classification model using a patient's breathing sounds. METHODS: Breathing sounds were recorded from 83 subjects during a PSG test. There was no exclusive experimental protocol or additional recording instruments use throughout the sound recording procedure. Based on the Apnea-Hypopnea Index (AHI), which indicates the severity of sleep apnea, the subjects' sound data were divided into four-OSA severity classes. From the individual sound data, we proposed two novel methods which were not attempted in previous OSA severity classification studies. First, the total transition probability of approximated sound energy in time series, and second, the statistical properties derived from the dimension-reduced cyclic spectral density. In addition, feature selection was conducted to achieve better results with a more relevant subset of features. Then, the classification model was trained using support vector machines and evaluated using leave-one-out cross-validation. RESULTS: The overall results show that our classification model is better than existing multiple OSA severity classification method using breathing sounds. The proposed method demonstrated 79.52% accuracy for the four-class classification task. Additionally, it demonstrated 98.0% sensitivity, 75.0% specificity, and 92.78% accuracy for OSA subject detection classification with AHI threshold 5. CONCLUSIONS: The results show that our proposed method can be used as part of an OSA screening test, which can provide the subject with detailed OSA severity results from only breathing sounds.