Modeling Obstructive Sleep Apnea using Dynamical Phase Transition

Sleep is an essential process to maintain and improve human activities, while many details related to sleep are still not well understood. Decreased or fragmented sleep is a health risk that might result in heart disease or diabetes on one hand and degradation of consciousness and cognition on the other hand. Sleep fragmentation is a phenomenon in which an individual's sleep is intermittently disrupted by arousal caused by external factors (noise) or internal factors (apnea) although sleep deprivation does not completely occur. Computational modeling is a suitable framework for understanding complex biological mechanisms. In this paper, the fundamental phenomena underlying the sleep-wake transition was reviewed and simulated. The dynamical behavior of model was then investigated and afterwards the factors that might cause obstructive sleep apnea were implemented and evaluated. The model includes two main neuronal populations: the ascending arousal system in the brain stem that is responsible for awakening and a neuronal population in the hypothalamus, called VLPO, which mediates sleep. These populations have mutual inhibition on each other causing a flip-flop or switching behavior between sleep and wake. The results of modeling in this paper showed hysteresis in the sleep-wake cycle, the size of which is affected by factors causing arousal. In OSA, intermittent and unstable transitions as well as the shrinking of bistable zone is expected. The model could reproduce some experimental results related to obstructive apneas.