Loss of consciousness, drowsiness, loss of awareness and in general oscillations in arousal are a fluctuating aspect of brain and cognitive function of the highest importance to health and to cognitive attainment [1]. Their intrinsic dynamics are at the root of healthy and clinical dynamics. In fact, in the absence of a gold standard for conscious response measurement, there are composite of approaches that rather diversify interpretability and lower the impact of intervention at the bedside [2]. Hence, the functional dynamics of sleep and their impact in health and cognition remain in coarse resolution, while the effects of traumatic brain injuries and their progress are still largely difficult to forecast and assign treatment [3]. To tackle this lack of specificity and effect, a phase-state reconstruction and representation learning approach is implemented, to build larger scale model of temporal dynamics in a typical duration of an fMRI experiment, in this case of a sleep deprivation resting-state scan from the Stockholm Sleepy Brain project [4] with the aim to later improve TBI evaluation. Moreover, several temporal regime approaches will be tested and a generative repertoire of pattern-states be organized for better experimentation and possible clinical interest [5,6]. 1. Demertzi A, Tagliazucchi E, Dehaene S, Deco G, Barttfeld P, Raimondo F, et al. Human consciousness is supported by dynamic complex patterns of brain signal coordination. Science Advances. 2019 Feb 1;5(2):eaat7603. 2. Viola-Saltzman M, Watson NF. Traumatic Brain Injury and Sleep Disorders. Neurol Clin. 2012 Nov;30(4):1299–312. 3. Skibsted AP, Amiri M, Fisher PM, Sidaros A, Hribljan MC, Larsen VA, et al. Consciousness in Neurocritical Care Cohort Study Using fMRI and EEG (CONNECT-ME): Protocol for a Longitudinal Prospective Study and a Tertiary Clinical Care Service. Front Neurol. 2018. 4. Nilsonne G, Tamm S, Lavebratt C, Liu JJ, Månsson KNT, Sundelin T, et al. A multimodal brain imaging dataset on sleep deprivation in young and old humans: The Stockholm Sleepy Brain Study. :27. 5. Kim JH, Zhang Y, Han K, Wen Z, Choi M, Liu Z. Representation learning of resting state fMRI with variational autoencoder. NeuroImage. 2021 Nov 1;241:118423. 6. Simidjievski N, Bodnar C, Tariq I, Scherer P, Andres Terre H, Shams Z, et al. Variational Autoencoders for Cancer Data Integration: Design Principles and Computational Practice. Frontiers in Genetics.