Detection of sleep apnea using only inertial measurement unit signals from apple watch: a pilot-study with machine learning approach

Sleep And Breathing, Год журнала: 2025, Номер 29(1)

Опубликована: Фев. 1, 2025

Abstract Purpose Despite increased awareness of sleep hygiene, over 80% apnea cases remain undiagnosed, underscoring the need for accessible screening methods. This study presents a method detecting using data from Apple Watch’s inertial measurement unit (IMU). Methods An algorithm was developed to extract seismocardiographic and respiratory signals IMU data, analyzing features such as breathing heart rate variability, dips, body movements. In cohort 61 adults undergoing polysomnography, we analyzed 52,337 30-second epochs, with 12,373 (23.6%) identified apnea/hypopnea episodes. Machine learning models five classifiers (Logistic Regression, Random Forest, Gradient Boosting, k-Nearest Neighbors, Multi-layer Perceptron) were trained on 41 subjects validated 20 subjects. Results The Forest classifier performed best in per-epoch event detection, achieving an AUC 0.827 F1 score 0.572 training group, 0.831 0.602 test group. model’s per-subject predictions strongly correlated apnea-hypopnea index (AHI) polysomnography ( r = 0.93) AHI ≥ 15 100% sensitivity 90% specificity. Conclusion Utilizing widespread availability Watch low power requirements IMU, this approach has potential significantly improve accessibility.

Язык: Английский

---

Review and perspective on sleep-disordered breathing research and translation to clinics

Sleep Medicine Reviews, Год журнала: 2023, Номер 73, С. 101874 - 101874

Опубликована: Ноя. 25, 2023

Sleep-disordered breathing, ranging from habitual snoring to severe obstructive sleep apnea, is a prevalent public health issue. Despite rising interest in and awareness of disorders, research diagnostic practices still rely on outdated metrics laborious methods reducing the capacity preventing timely diagnosis treatment. Consequently, significant portion individuals affected by sleep-disordered breathing remain undiagnosed or are misdiagnosed. Taking advantage state-of-the-art scientific, technological, computational advances could be an effective way optimize treatment pathways. We discuss multidisciplinary research, review shortcomings current SDB management adult populations, provide possible future directions. critically opportunities for modern data analysis machine learning combine multimodal information, perspective pitfalls big analysis, approaches developing strategies that overcome limitations. argue large-scale collaborative efforts based clinical, technical knowledge rigorous clinical validation implementation outcomes practice needed move forward, thus increasing quality diagnostics

Язык: Английский

---

Screening for obstructive sleep apnea hypopnea using sleep breathing sounds based on the PSG-audio dataset

Biomedical Signal Processing and Control, Год журнала: 2025, Номер 103, С. 107472 - 107472

Опубликована: Янв. 1, 2025

Язык: Английский

---

Perspectives Regarding Consumer Sleep Technology and Barriers to its Use or Adoption Among Adults in the United States

Sleep Medicine, Год журнала: 2025, Номер 128, С. 165 - 173

Опубликована: Фев. 5, 2025

Язык: Английский

---

Management of Adult Obstructive Sleep Apnoea: Many Questions, Not Enough Answers!

Journal of Sleep Research, Год журнала: 2025, Номер unknown

Опубликована: Март 24, 2025

Obstructive sleep apnoea (OSA) conveys a substantial global public burden due to its high prevalence and causative relationship with cardiometabolic diseases. The current diagnostic reliance on the apnoea/hypopnoea index (AHI) is insufficient address complex, multifaceted condition, revision of standard criteria urgently needed. Together better understanding clinical, pathophysiological, phenotypic characteristics, this will pave way personalised, holistic treatment approaches.

Язык: Английский

---

Detection of sleep apnea using smartphone-embedded inertial measurement unit

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 28, 2025

Abstract We previously demonstrated that sleep apnea (SA) can be detected using acceleration and gyroscope signals from smartwatches. This study investigated whether an inertial measurement unit (IMU) embedded in non-wristwatch devices, such as smartphones, also detect SA when worn during sleep. During polysomnography (PSG), subjects wore IMU-embedded GPS device (Amue Link ® ) and/or smartphones (Xperia or iPhone on their abdomen. Triaxial were recorded overnight. Data split into training test groups (2:1) for each device. An algorithm was developed the to extract respiratory movements (0.13–0.70 Hz) events, which validated groups. IMU-derived events showed breath-by-breath concordance with PSG apnea-hypopnea yielding F1 scores of 0.786, 0.821, 0.796, respectively. Regression model derived IMU correlated AHI ( r = 0.90, 0.93, 0.96), limits agreement -16.7 25.9, -17.4 22.5, − 18.4 20.5. Using cutoff values groups, moderate-to-severe (AHI ≥ 15) identified AUCs 0.95, 0.98, 0.94 0.89, 0.96, 0.92, IMUs including quantitatively

Язык: Английский

---

AI-Driven Detection of Obstructive Sleep Apnea Using Dual-Branch CNN and Machine Learning Models

Biomedicines, Год журнала: 2025, Номер 13(5), С. 1090 - 1090

Опубликована: Апрель 30, 2025

Background/Objectives: The purpose of this research is to compare and contrast the application machine learning deep methodologies such as a dual-branch convolutional neural network (CNN) model for detecting obstructive sleep apnea (OSA) from electrocardiogram (ECG) data. Methods: This approach solves limitations conventional polysomnography (PSG) presents non-invasive method OSA in its early stages with help AI. Results: shows that both CNN models can identify ECG signals. achieves validation test accuracy about 93% 94%, respectively, whereas 94% accuracy. Furthermore, obtains ROC AUC score 0.99, meaning it better at distinguishing between non-apnea cases. Conclusions: results show models, especially CNN, are effective classification than traditional methods. In addition, our proposed has potential be used reliable, accurate detection even current state-of-the-art advanced

Язык: Английский

---

Validation of Downloadable Mobile Snore Applications by Polysomnography (PSG)

Nature and Science of Sleep, Год журнала: 2024, Номер Volume 16, С. 489 - 501

Опубликована: Май 1, 2024

Purpose: Obstructive sleep apnea (OSA) is a common breathing disorder during that associated with symptoms such as snoring, excessive daytime sleepiness, and interruptions. Polysomnography (PSG) the most reliable diagnostic test for OSA; however, its high cost lengthy testing duration make it difficult to access many patients. With availability of free snore applications home-monitoring, this study aimed validate top three ranked applications, namely SnoreLab (SL), Anti Snore Solution (ASS), Sleep Cycle Alarm (SCA), using PSG. Patients Methods: Sixty participants underwent an overnight PSG while simultaneously identical smartphones tested apps gather snoring data. Results: The discovered all were significantly correlated total recording time counts PSG, ASS showing good agreement counts. Furthermore, Score, Time Snoring SL, Quality SCA had significant correlation natural logarithm hypopnea index (lnAHI) Score SL shown be useful evaluating severity pre-diagnosing or predicting OSA above moderate levels. Conclusion: These findings suggest some parameters can employed monitor progress, future research could involve adjusted algorithms larger-scale studies further authenticate these downloadable applications. Keywords: polysomnography, smartphone apps, obstructive apnea,

Язык: Английский

---

Engineering Triboelectric Paper for Energy Harvesting and Smart Sensing

Advanced Materials, Год журнала: 2024, Номер unknown

Опубликована: Дек. 17, 2024

Triboelectric nanogenerators (TENGs) represent a promising technology for energy harvesting and self-powered sensing with wide range of applications. Despite their potential, challenges such as the need cost-effective, large-area electrodes engineering sustainable triboelectric materials remain, especially given impending restrictions on single-use plastics in Europe. To address these challenges, nano-graphite-coated paper is presented high-performance alternative layers. Moreover, this material, which can be produced an industrial scale, offers viable replacement metal electrodes. The combination nano-graphite paper, its large contact area inherent surface roughness, enables ultra-high power densities exceeding 14 kW m

Язык: Английский

---

Application of p and n-Type Silicon Nanowires as Human Respiratory Sensing Device

Sensors, Год журнала: 2023, Номер 23(24), С. 9901 - 9901

Опубликована: Дек. 18, 2023

Accurate and fast breath monitoring is of great importance for various healthcare applications, example, medical diagnoses, studying sleep apnea, early detection physiological disorders. Devices meant such applications tend to be uncomfortable the subject (patient) pricey. Therefore, there a need cost-effective, lightweight, small-dimensional, non-invasive device whose presence does not interfere with observed signals. This paper reports on fabrication highly sensitive human respiratory sensor based silicon nanowires (SiNWs) fabricated by top-down method metal-assisted chemical-etching (MACE). Besides other important factors, reducing final cost paramount importance. One factors that increases price sensors using gold (Au) electrodes. Herein, we investigate sensor's response aluminum (Al) electrodes as cost-effective alternative, considering fact electrode's work function crucial in electronic design, impacting properties electron transport efficiency at electrode-semiconductor interface. Therefore comparison made between SiNWs from both p-type n-type effect dopant electrode type sensing functionality. A distinct directional variation was sample's Au Al Finally, performing qualitative study revealed electrical resistance across renders greater sensitivity than dry air pressure. No definitive research demonstrating mechanism behind these effects exists, thus prompting our underlying process.

Язык: Английский

---