Abstracts

Rationale: Direct stimulation of the human brain can elicit sensory and motor perceptions as well as recall of memories. Stimulating higher order association areas of the temporal cortex in particular was shown to activate multi-modal memory representations of past experiences (Penfield and Perot 1963, Brain 86). Hypothetically, this could be utilized to develop new treatments for cognitive deficits in patients suffering from epilepsy and other brain disorders . Most of the recent attempts at memory enhancement in human patients have focused on the hippocampus and the associated medial temporal lobe structures reporting mixed results in small subject groups (Kim et al. 2016, Neurobiology of Learning and Memory, in press). Here we used a large dataset of intracranial recordings with stimulation in epilepsy patients to investigate the effect of targeted local stimulation on memory performance across a range of brain structures. Methods: In total, 40 patients implanted with intracranial electrodes for seizure monitoring were stimulated during encoding of word lists for subsequent recall in two verbal memory tasks. In each task subjects remembered a list of 12 words for subsequent free recall following a math distractor problem. 50Hz continuous bipolar stimulation was delivered during epochs of word presentation through a pair of neighboring electrode contacts selected based on their anatomical localization and neurophysiological patterns of activity established from preceding passive task recordings. 22 distinct brain structures were stimulated to assess the effect on memory recall. Results: We report memory enhancement in two out of two cases of stimulation in the left posterior middle temporal gyrus, which resulted in significantly increased number of remembered words on stimulated versus non-stimulated lists (p < 0.05, permutation test) with subjective experience of improved remembering of words in one of the patients. The effect of stimulation was correlated with univariate changes in spectral power, coherence and phase synchrony, as well as by a multi-variate classifier analysis of spectral power changes characterizing successful word recall. There was no positive effect found in any other of the structures tested in this study, which included areas of the prefrontal cortex, hippocampus and the associated medial temporal neocortex. Conclusions: To our knowledge, this is the first report of memory enhancement induced by superficial stimulation of the temporal neocortex. Our results support intelligent biomarker-driven mapping and targeting of specific cortical networks for development of novel neurotechnologies to modulate memory functions and treat cognitive deficits in humans. Funding: This work was supported by the DARPA Restoring Active Memory (RAM) program (Cooperative Agreement N66001-14-2-4032).