Insomnia is increasingly recognized not only as a sleep disorder but as a complex neurocognitive dysregulation involving persistent alterations in thought patterns, emotional processing, and awareness stability. Traditional clinical models emphasize behavioral and physiological dimensions; however, recent advances in computational psychiatry suggest that insomnia can be effectively conceptualized as a dynamic system of disrupted cognitive states. This study proposes an advanced computational architecture for modeling cognitive state transitions in insomnia, with a focus on thought intrusion, attentional bias, and awareness fragmentation.

Drawing on formal psychopathology modeling frameworks (Haslbeck et al., 2022) and neurocomputational causal modeling approaches (Pereira et al., 2021), the proposed architecture integrates multi-layered cognitive variables to simulate insomnia-related cognitive instability. The model incorporates repetitive negative thinking dynamics (Lancee et al., 2015), attentional bias mechanisms (Milkins et al., 2016), and cognitive behavioral therapy response pathways (Trauer et al., 2015).

Additionally, this research situates computational insomnia modeling within broader machine learning and deep learning paradigms, including adaptive learning systems used in cybersecurity and behavioral prediction domains (Akram et al., 2024; Cheng et al., 2024). The findings highlight the potential of computational architectures in predicting insomnia severity, mapping cognitive disruptions, and optimizing therapeutic interventions.

The study concludes that insomnia can be effectively represented as a nonlinear cognitive-state transition system, where awareness instability emerges from interacting cognitive, emotional, and attentional subsystems.