Long-term Coding of Associations in the Human Medial Temporal Lobe
thesisposted on 23.01.2018, 15:34 by Emanuela De Falco
This PhD thesis aims to investigate the role of human medial temporal lobe (MTL) neurons in the encoding of associations between items and the characteristics of single neuron activity in response to related items. The invaluable opportunity to record single neuron activity occurs when patients suffering from refractory epilepsy have to be implanted with intracranial depth electrodes to treat their clinical condition. It has been long recognised that the MTL plays a critical role for declarative memory functions. MTL neurons have been shown to change their tuning during associative learning tasks. However, it is still not clear whether their involvement is confined to the task execution or goes beyond it. To address this issue, we studied the responses of MTL neurons in neurosurgical patients to known concepts (people and places), in conjunction with two different metrics measuring the degree of association between items (one metric based on the patients' evaluations and the other based on web searches). We found that whenever MTL neurons responded to more than one concept, these concepts were typically related, therefore providing evidence for a long-term involvement of MTL neurons in the representation of durable associations, which is essential to declarative memory functions. We also analysed the differences between spiking responses elicited by different stimuli in a single neuron, and how these differences related to the degree of association between stimuli. We found that, in general, MTL neurons exhibit a similar neural activity in response to different stimuli, and that eventual differences in the responses are smaller the more two stimuli are associated to each other. Our results support the idea that a process of unitisation of neural responses occurs in the MTL, and that information about the stimulus identity is not encoded in individual neurons, but rather at the neural assembly level.