A hybrid approach for continuous detection of sleep‐wakefulness fluctuations: validation in patients with Cheyne–Stokes respiration

Abstract

SUMMARYFluctuations in sleep ⁄ wake state have been proposed as an important mechanism contributing to the development of oscillatory breathing patterns, including Cheyne-Stokes respiration in patients with heart failure. In order to properly assess the interactions between changes in state and changes in ventilatory parameters, a methodology capable of continuously and reliably detecting state transitions is needed. Traditional fixed-epoch analysis of polysomnographic recordings is not suitable for this purpose. Moreover, visual identification of changes in the dominant electroencephalogram activity at the transition from wakefulness to sleep and vice versa is often very subjective. We have therefore developed a hybrid approach -including both visual scoring and computer-based procedures -for continuous analysis of state transitions from polysomnographic recordings, specifically tailored for fluctuations between wakefulness and non-rapid eye movement-1 and -2 sleep. The overall analysis process comprises three major phases: (1) manual identification of relevant electroencephalogram ⁄ electrooculogram features and events, including a sample of unequivocal alpha and theta-delta activity; (2) automatic statistical discrimination of dominant electroencephalogram activity; and (3) state classification (wakefulness, non-rapid eye movement-1 and -2). The latter is carried out by merging information from visual scoring with the output of the discriminator. Validation has been carried out in 16 patients with heart failure during daytime Cheyne-Stokes respiration, using a training and testing set of electroencephalogram polysomnograms. The statistical discriminator correctly classified 99.1 ± 1.4% and 99.2 ± 1.1% of unequivocal alpha and theta-delta activity. This approach has therefore the potential to be used to reliably measure the incidence and location of sleep-wake transitions during abnormal breathing patterns, as well as their temporal relationship with major ventilatory events.

Respiratory and autonomic dysfunction in children with autism spectrum disorders

“…CSR is a type of periodic breathing seen in individuals with neurological disorders (brainstem lesions, multiple system atrophy and others) [23] and is also seen in people with heart failure [24]. CSR could occur during wakefulness, drowsiness or in sleep.…”

Sleep–wake fluctuations and respiratory events during Cheyne–Stokes respiration in patients with heart failure

“…Occasional W → NREM → W transitions were observed in many patients during the crescendo or decrescendo phase of hyperpnoea, and were likely a manifestation of unstable transitional patterns in the EEG (Ogilvie, ; Pinna et al ., ; Santamaria and Chiappa, ). Interestingly, transitions occurring during the crescendo phase were often associated with a depression of the concomitant hyperpnoea, thus suggesting that the lack of the extra drive to breathe associated with wakefulness led to a reduction of the cyclic hyperpnoea.…”

“…We used a method that includes both visual and computer‐based analysis of polysomnographic recordings (Pinna et al ., ). Details are provided in the Supporting Information.…”

“…The purpose of this study was to test this hypothesis using a novel methodology for continuous detection of fluctuations between wakefulness and NREM 1–2 sleep (Pinna et al ., ). This approach allowed us to detect state transitions with a time resolution of 0.25 s, rather than every 30 s as in conventional polysomnography, and to consider state changes as short as 2 s. Moreover, it allowed us to identify the occurrence of wake–sleep–wake transitions objectively, and to accurately assess their temporal relationship with apnoeic events.…”

A signal demodulation-based method for the early detection of Cheyne-Stokes respiration

“…Some methods have been proposed to quantify the amplitude of the oscillations. For example, a spectral decomposition algorithm of the instantaneous minute ventilation is proposed in [15][16][17] where periodic breathing has to be previously detected to be quantified. A method based on a standard amplitude demodulation scheme based on filters is presented in [18].…”