Visual and automatic classification of the cyclic alternating pattern in electroencephalography during sleep

Abstract

Cyclic alternating pattern (CAP) is a neurophysiological pattern that can be visually scored by international criteria. The aim of this study was to verify the feasibility of visual CAP scoring using only one channel of sleep electroencephalogram (EEG) to evaluate the inter-scorer agreement in a variety of recordings, and to compare agreement between visual scoring and automatic scoring systems. Sixteen hours of single-channel European data format recordings from four different sleep laboratories with either C4-A1 or C3-A2 channels and with different sampling frequencies were used in this study. Seven independent scorers applied visual scoring according to international criteria. Two automatic blind scorings were also evaluated. Event-based inter-scorer agreement analysis was performed. The pairwise inter-scorer agreement (PWISA) was between 55.5 and 84.3%. The average PWISA was above 60% for all scorers and the global average was 69.9%. Automatic scoring systems showed similar results to those of visual scoring. The study showed that CAP could be scored using only one EEG channel. Therefore, CAP scoring might also be integrated in sleep scoring features and automatic scoring systems having similar performances to visual sleep scoring systems.

A Portable Wireless Device for Cyclic Alternating Pattern Estimation from an EEG Monopolar Derivation

“…It was verified that the majority of the misclassifications occur in the CAP events boundary, agreeing with the conclusions reported by Largo et al [ 15 ] about the importance of accurately measuring these boundaries. This issue was mitigated by the post-processing procedure employed that provided, on average, an increase of 2% in the Acc .…”

“…Consequently, the mutual agreement among two physicians, examining the same EEG signals, ranges from 69% to 77.5% [ 14 ]. The agreement gets closer to the lower bound as more physicians are considered in the agreement analysis, as reported by Largo et al [ 15 ], where the global average of the pairwise inter-scorer agreement of seven experts was 69.9%. Therefore, an automatic scorer algorithm implemented in a home monitoring device (HMD) is desired to address this issue.…”

“…It was verified that CAP has a strong temporal dependency, as indicated by Tezano et al [ 5 ], that allowed the LSTM based classifier to achieve an average performance that is in the higher bound of the experts’ agreement range indicated by Rosa et al [ 14 ] and significantly higher than the inter-scorer agreement of multiple experts reported by Largo et al [ 15 ]. It is also a relevant fact that the algorithm does not need to remove the REM periods (common practice for CAP scoring), making the implementation more suitable for clinical analysis.…”

“…However, Largo et al [ 15 ] have verified that the global average of the pairwise inter-scorer agreement from seven specialists performing visual scoring was 69.9%, suggesting that the agreement gets closer to the lower bound (69%) as the number of experts that are performing the scoring increases. Hence, the method we developed is, at least, as good as a group of specialists analyzing the same signals and the implementation in an HMD could possibly be used as a diagnostic tool, potentiating the detection and treatment of sleep quality deficits.…”

Matrix of Lags: A tool for analysis of multiple dependent time series applied for CAP scoring

“…The highest achieved accuracy, for the CAP cycle detection using the ECG signal (developed method for indirect estimation of the CAP cycles), was 77%, a value that is in the specialist agreement range (69% to 78%) analyzing the same EEG signal [23] . This agreement gets closer to the lower bound when the number of physicians performing the scoring increases as verified by Largo et al [54] where seven independent scorers achieved an inter-scorer agreement of 69.9%. Therefore, the developed method is as good as the specialists scoring the CAP cycles from the EEG signal using the ECG signal, supporting the viability of the proposed method that performs an indirect classification using a simpler to self-assembly sensor.…”