0112 Classification of reconstructed depth profiles shows Global and non-Global slow oscillations differentiate in the hippocampus and thalamus

Abstract Introduction Sleep slow oscillations (SOs, 0.5-1.5 Hz) can be classified on the scalp as Global, Local or Frontal, where Global SOs are found in most electrodes within a short time delay and gate long-range information flow during NREM sleep. In this study, we estimate current density brain that generates SO, to evaluate which sub-cortical structures involved SO dynamics. We then train multiple machine learning algorithms distinguish between other types, probe variance of Global/non-Global profiles across subjects.Sleep subjects. Methods 32 volunteers (18 females) slept lab with polysomnography including 24 head EEG channels; their sleep was scored according AASM criteria. were algorithmically detected at each channel non-Global using our method (Malerba et al., 2019). The depth profile reconstructed source estimation (in Brainstorm followed by sLORETA), standardized model 17 regions. Each embedded matrix averaging region three 200ms-long bins: before, after trough. Thirty classifiers applied dataset, leveraging Matlab’s supervised application. compared accuracy subjects identified best-performing dataset size. used univariate feature selection quantify relevance region-time pair successful classification. Results have higher subjects, improves when data is sampled rather than individuals. Ensemble subspace methods reached highest (98.5%). Feature selectivity cortical, hippocampal, thalamic currents trough relevant for classifications. Conclusion introduce an analytical framework enabling study profiles, evolution, matrices. predominant differentiation supports potential functional difference these types. Support (If Any) NIH grant (R01 AG046646) S.C.M.