Key result
ASLT combined with Inception-v3 detects sleep disorders with ~98% accuracy.
Why the study?
Assumptions regarding sleep microstructure differences produce complexity in CVD analysis, motivating the examination of new labels and energy-based spectrum interpretations of cyclic alternating pattern (CAP) events from ECG signals to characterize sleep disorders.
Does an adaptive superresolution transform (ASLT) with an Inception-v3 deep neural network improve sleep disorder detection and heart rate estimation compared to STFT and CWT in the CAP sleep dataset?
Does an adaptive superresolution transform (ASLT) with an Inception-v3 deep neural network improve sleep disorder detection and heart rate estimation compared to STFT and CWT in the CAP sleep dataset?
Combining adaptive superresolution transform (ASLT) with an Inception-v3 deep neural network on ECG signals achieves 98.2% accuracy in detecting various sleep disorders and improves heart rate estimation.
Answers are grounded in this paper and cite their sources.
May aid ECG-based sleep research; leaves open prospective validation before clinical adoption.
This study aims to examine the new labels for the significant characterization of sleep disorder, in the context of sleep microstructure, as revealed by cyclic alternating pattern (CAP). The observation of sleep disorders through ceaseless recording and monitoring of electrocardiogram (ECG) signal assist to take protective care upon cardiovascular diseases (CVDs). Differences between normal subject and sleep disordered subject are scrutinized by estimating the heart rate. The variation of CAP events is then studied in distinct scales by the use of wavelet analysis. The essence of CAP events is considered in terms of electrocardiography (ECG) complexity analysis for various type of events. Few assumptions regarding the microstructure difference in sleep disorders including quality sleep, asymmetric activation-deactivation patterns, high duration of desynchronization phases, produces complexity in CVD analysis. This study expands the existing holdings between the increased CAP rate and the instable characteristics of sleep, highlight the differences in sleep disorders with categorization analysis. This study creates a new aspect for the interpretation of the CAP events through energy-based spectrum. The detection of sleep disorders including insomnia, narcolepsy, nocturnal frontal lobe epilepsy (NFLE), rapid eye movement sleep behavior disorder (RBD), and sleep disordered breathing (SDB) with normal is performed. The proposed methodology employed adaptive superresolution transform (ASLT) with Inception-v3 deep neural network attained an accuracy of 98.2% using CAP sleep dataset for sleep disorder detection. The proposed methodology outperforms the state-of-the-art methods namely short time Fourier transform (STFT), and continuous wavelet transform (CWT) in heart rate estimation.
No takes yet. Share an insight, caveat, or question.
Singhal et al. (2025) studied Sleep disorders (insomnia, narcolepsy, nocturnal frontal lobe epilepsy, rapid eye movement sleep behavior disorder, and sleep disordered breathing). Adaptive superresolution transform (ASLT) with Inception-v3 deep neural network vs. Short time Fourier transform (STFT) and continuous wavelet transform (CWT) was evaluated on Sleep disorder detection accuracy. Adaptive superresolution transform (ASLT) with Inception-v3 deep neural network attained an accuracy of 98.2% using the CAP sleep dataset for sleep disorder detection.
Synapse has enriched 10 related papers. Consider them for comparative context: