Assessment of Cyclic Alternating Pattern of sleep during late pregnancy

Cyclic Alternating Pattern (CAP) is an indicator of sleep instability often associated with various sleep disorders. CAP is characterized by transient changes in cerebral electrical activity, known as A-phases, that occur during sleep. These events are a fundamental component of sleep microstructure and serve as a metric for assessing sleep quality. This research aims to investigate the sleep patterns of pregnant women with good sleep quality in the third trimester to assess sleep behavior from a CAP perspective. An ensemble of deep learning models was developed and trained on electroencephalography (EEG) data from the Physionet CAP database and applied to automatically detect A-phases in EEG data from pregnant women. Statistical analyses were performed using various CAP indices calculated from the detected A-phases in pregnant and non-pregnant women who served as a control group. The results show that pregnant women have a reduction of approximately 10% in the CAP rate index compared to non-pregnant group. This reduction seems to be independent of the non-rapid eye movement (NREM) duration since it is similar in both groups. This finding is interesting because the CAP rate generally represents an increase in the pathological state. The results of this study show that pregnant women have a unique behavior of CAP, which could be an additional indicator of sleep quality in pregnancy. Clinical relevance - In the context of obstetrics and gynecology focused on pregnant women, CAP has not been evaluated, so this work opens the door to include new indicators such as sleep instability, which could provide relationship with the status of pregnant women.