Neuromodulation of human memory: From single neuron recordings to clinical intervention

Présentation de la conférence

Loss of memory is one of the most devastating afflictions of the human condition. Present in ever increasing numbers of neurological patients it casts an ominous shadow on the social horizon. The medial temporal lobe, often affected in these disorders, is central to the transformation of percepts into lasting memories that can be consciously recollected in the future. Yet the neuronal code underlying this transformation in humans remains unclear. While animal models provide an important basis for the study of memory, there is no substitute for the study of human subjects who can declare their percepts and memories. Neurological patients implanted with intracranial depth electrodes for clinical reasons, present a rare opportunity to study human memory at the single neuron level. Recent evidence demonstrates several striking properties of neurons in hippocampus and entorhinal cortex associated with encoding of information, imagery, free recall and spatial memory. These properties include high degree of response specificity to stimulus identity, a striking invariance with respect to physical features and modality of stimuli, correlation with conscious perception and degradation during sleep deprivation, reinstatement during visual recall and imagery, and dynamic incorporation of associated stimuli by single neurons. These studies suggest sparse coding at the interface of conscious perception and declarative memory, by small ensembles of “concept cells” involved in the encoding and reinstatement of abstract information. Furthermore, electrical stimulation applied at the entorhinal region during encoding of information enhances future memory of this information. These findings may provide a basis for development of closed -loop cognitive neuroprosthetic devices to assist memory function in neurological patients. Such devices may be based on site-specific complex stimulation patterns coupled to information processing stages, and operating in on-demand fashion with neuronal signal feedback.