Abstracts

FMRI MEMORY ASSESSMENT IN HEALTHY SUBJECTS: A NEW APPROACH TO VIEW LATERALIZATION DATA ON AN INDIVIDUAL LEVEL

Abstract number : 1.145
Submission category : 5. Human Imaging
Year : 2009
Submission ID : 9528
Source : www.aesnet.org
Presentation date : 12/4/2009 12:00:00 AM
Published date : Aug 26, 2009, 08:12 AM

Authors :
Maria Strandberg, C. Elfgren, I. Rorsman and K. K ll n

Rationale: Functional neuroimaging makes it possible to study the neural correlates of both encoding and retrieval processes of episodic memory in the medial temporal lobes (MTL). The aim of our study was to develop an fMRI paradigm for assessment of memory capacity and laterality to be used in epilepsy surgery units where reliable individual memory evaluation is an important prerequisite. Methods: 15 right-handed healthy volunteers underwent fMRI in a 3T MRI unit performing: (A) a verbal encoding memory paradigm with deep versus shallow encoding with a subsequent recognition test; and (B) a visuo-spatial recall paradigm. We used a semi-random design, modelling both event-related and block effects. In addition to whole-brain analysis, 2 masks including anatomic regions of interest (ROI) were constructed: (I) an MTL ROI; and (II) a Broca ROI. Laterality indices (LI) and curves were calculated using MATLAB and a toolbox running within the SPM5 environment. The toolbox produces a numerical mean LI value with an attached confidence and a curve where LI is presented as a function of increasing fMRI activation threshold for each individual. Mean LI is a mean of all values that constitutes the curve. Results: Results from the MTL ROI and verbal paradigm analysis (I.A) showed left-lateralized activity in 12 of the 15 subjects, LI group median 0.24 (range -0.26 - 0.64). Main regions of strong activation were the left prefrontal cortex and the left MTL. Hemispheric language dominance (II.A) was lateralized to the left for all but one subject, LI group median 0.53 (range -0.31 - 0.88). Visuospatial MTL (I.B) activation showed no lateralization with LI close to zero, group median 0.018 (range -0.36 - 0.35). The peak maximum for the whole brain was in the right middle temporal gyrus, but local maxima was seen in the left fusiform gyrus and the parahippocampal cortices bilaterally for the this task. The laterality indices changed with increasing threshold for all subjects for both the deep-shallow contrast and the active-rest contrast during the visuospatial task. For each individual three different LI curves were produced and these single subject curves yielded additional information as they facilitated trend assessment simultaneously as determination of hemisphere dominance was accomplished. Conclusions: Our results were concordant with previously reported group data and demonstrated the problems with threshold-dependent activation patterns and large within group variability. Our paradigm comprises two different tasks testing both verbal and visuospatial memory. We present a new approach to achieve a comprehensive view of single subject memory performance where a lateralization curve, an overall LI mean with an attached confidence interval, can be combined with morphological images and behavioural data. The laterality curve approach is applicable on both patients and on healthy subjects and presents overall clinical relevant information on a single subject level.
Neuroimaging