AUTHOR=Sanchez Panchuelo Rosa M. , Flintham Robert , Wesolowski Roman , Jalali Roya , Herbert Jane , Samarasekera Shanika , Bagshaw Andrew P. , Chelvarajah Ramesh , Davies Nigel P. , Sawlani Vijay TITLE=Test-retest reliability of the “Home Town Walk” fMRI paradigm for memory activation and lateralization in the pre-surgical evaluation of patients with temporal lobe epilepsy JOURNAL=Frontiers in Neurology VOLUME=15 YEAR=2024 URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2024.1419047 DOI=10.3389/fneur.2024.1419047 ISSN=1664-2295 ABSTRACT=Introduction

Functional magnetic resonance imaging (fMRI) can be used to assess language and memory function as part of pre-surgical decision making in refractory epilepsy. Although language paradigms are well established, memory paradigms are not widely used in clinical practice due to a lack of evidence for robust and reliable methods. Here, we aim to investigate the clinical utility of the Home Town Walk (HTW) paradigm for personalized treatment decisions in medial temporal lobe epilepsy.

Methods

A cohort of 123 consecutive patients having HTW-fMRI as part of routine MRI scans over a 7.5 year period were included in this retrospective study. Of these, 111 patients underwent repeated HTW-fMRI in two scanning sessions one to three days apart. fMRI analysis was performed at the time of the scans using clinically approved software and retrospectively validated using FSL. We assessed the test–retest within subject reliability of activations within the posterior parahippocampal gyri (pPHG) at the individual subject level.

Results and discussion

Activations within the pPHG region were observed for 101 patients (91%) in at least one of the fMRI sessions and for 88 patients (79%) in both fMRI sessions, with 82 patients showing overlapping unilateral or bilateral activations and 8 further patients showing overlapping activations in one of the hemispheres but not the other. Reproducibility was evaluated using metrics based on the concordance ratios for size (Rsize) and location (Roverlap) within the pPHG region, as well as the lateralization index (LI) metric to reflect the asymmetry of hemispheric activations, which is of crucial relevance to inform surgery. Test–retest reliability of visuospatial memory LIs, assessed by an intra-class correlation coefficient (ICC) yielded a value of 0.76, indicating excellent between session stability of memory lateralization.

Conclusion

The HTW-fMRI paradigm shows reproducible activations in the medial temporal lobes of individual epilepsy patients sufficient to consistently lateralize visuospatial memory function, demonstrating the clinical utility of HTW memory fMRI and its potential for application in the pre-surgical assessment of people with temporal lobe epilepsy.