Abstract
The hippocampus plays a crucial role in representing the temporal structure of events and potentially in detecting violations in expected sequences. This thesis investigated the impact of hippocampal dysfunction in Transient Global Amnesia (TGA) on the processing of sequence violations. TGA is characterised by a transient, but reversible phase of amnesia associated with temporary hippocampal disfunction, providing the unique opportunity to track associated impairments across recovery.
To date, there is a scarcity of studies recording brain activity during task performance in TGA patients. However, investigating brain responses when performing a task could reveal important details about the nature of impairments present during acute stages of TGA. With this, it could ultimately point to the specific involvements of CA1 in cognitive processes, which is a subregion of the hippocampus thought to be particularly affected by TGA-related effects. Using TGA as a disease model, this thesis aimed to explore whether violations of temporal sequences affect memory performance, and how hippocampal activity responds during and after the acute phase of TGA. To achieve this, task-related fMRI data of subjects with a TGA diagnosis were recorded at three distinct timepoints (acute phase, post-acute, and follow-up). The task paired a repetition encoding paradigm with a recognition memory test. Importantly, the encoding phase contained occasional temporal violations of expected sequence order, which was then followed by a subsequent target-foil discrimination task, including items from violated, as well as repeated sequences.
Behavioural results showed no substantial differences in memory performance between sequence violations and repetitions across all stages, nor between acute and post-acute/follow-up phase. Similarly, fMRI analysis revealed no significant hippocampal activation differences between sequence violations and repetitions during post-acute and follow-up sessions. However, when investigating the change of responses across sessions, a subset of voxels showed reversed sensitivity: they were more responsive to violation sequences compared to repeat sequences in the acute phase, while this was reversed for the post-acute/follow-up sessions. Because a spatially distinct set of voxels showed similar behaviour already at repetitions preceding the violation, this effect could not be attributed to the temporal violation. Further exploratory analyses tracking the change of responses across sessions revealed a cluster of voxels in the left hippocampus with significant changes to the last repeat of sequences when contrasted against baseline activity. While the responses were elevated in the post-acute/follow-up phase, they were much closer, or slightly lower than baseline activity in the acute phase. This may point to differences in how repetition affects BOLD activity between acute and recovered stages, or potentially indicate the presence of compensatory mechanisms. Further research is needed to clarify these findings.
While findings remain speculative, they underscore important methodological challenges given by task design sensitivity and sample size limitations that may have influenced the results. Although the study did not confirm the hypothesized memory benefit for sequence violations, it is a first step towards investigating task-based hippocampal processing during cognitive recovery in TGA. With initial findings and critical insight into task design, this thesis provides important recommendations, pointing out promising directions for future research.