K-complexes (KC) are events present in non-rapid eye movement (non-REM) sleep, which have cellular dynamics similar to slow waves and three distinguishing components: initial positive wave (P200), posterior negative wave (N500) and final positive wave (P900). They can be generated spontaneously in cortical networks but can also be induced by cortical, thalamic or sensory stimulation.

Sleep plays a fundamental role in memory consolidation, favoring the transfer of new information from the hippocampus to the neocortex and its cortico-cortical redistribution. There are no studies that directly link KCs with memory processes; however, reactivations during sleep, with cues associated with previously acquired information, induce evoked potentials similar to KCs. Therefore, KCs could be participating in the memory consolidation process during sleep and they are not being taken into account as a possible facilitating event of the hippocampal-cortical dialogue. Consequently, in order to study these events more deeply, we propose a Machine Learning based algorithm.