BACKGROUND AND PURPOSE:
Existing methods of resting state fMRI (rsfMRI) analysis are successful in localizing language function but are suboptimal for language lateralization. Most rsfMRI studies apply temporal correlation based analyses. The purpose of this study is to use a novel metric based on a 2D frequency domain power spectral analysis to determine hemispheric language lateralization from rsfMRI data.

METHODS:
We carried out a 2D Fourier analysis on three functional regions of interest (ROIs): 1) Broca’s Area (BA), 2) Hand-motor cortex (HM) and 3) Primary Visual Cortex (VC), on the rsfMRI data of 21 healthy, right-handed subjects. rsfMRI data analysis was performed in AFNI and Matlab. Spike removal, motion correction, linear trend removal and spatial smoothing were applied. Spontaneous low-frequency fluctuations (0.01-0.1Hz) were filtered to enable functional integration. We conducted a comparison of the resulting power spectral plots between the left and right BA, with the corresponding HM and VC included as controls, for each participant. A novel metric, termed the Power Lateralization Index (PLI), was used to quantify the extent to which spectral power differed between the left and right hemisphere. PLI was defined as:

PLI = (P_left - P_right) / (P_left + P_right)
where P_left = power metric for region in left hemisphere, and P_right = power metric for region in right hemisphere.

PLI for BA was subsequently compared to that of HM and VC. All reported z-statistics and p-values are from the Mann Whitney U-Test, to compare whether two sample means are significantly different. The Bonferroni correction was applied to counteract any potential issues with multiple comparisons.

RESULTS:
At the α=0.05 level, BA showed statistically significantly greater power in the left hemisphere compared to the right hemisphere (Z=2.69, P=0.0036), while HM (Z=1.03, P=0.1515) and VC (Z=-0.07, P=0.4681) showed no significant difference between the hemispheres. PLI for BA was statistically significantly greater than VC (Z=4.35, P<0.00001) and HM (Z=2.62, P=0.0044). Both sets of findings are consistent with left-lateralization of language function.
 
CONCLUSIONS:
Our method demonstrated left-lateralization of language in 21 right-handed healthy subjects based on rsfMRI data. Lateralization in the VC was symmetrical, as anticipated. HM showed a slight trend favoring left-lateralization that did not reach statistical significance, possibly due to right hand dominance. The outcome of this study supports the feasibility and validity of a 2D, ROI-based power-spectral analysis of rsfMRI data for language lateralization. We plan to validate this method in brain tumor patients.

REFERENCES:
1. Teghipco. NeuroImage: Clinical 12(2016):910-927
2. Yang. Neuroimage 36.1(2007):144-152

Frequency-Domain Resting State fMRI Analysis Demonstrates Language Lateralization in Healthy Controls

[Virtual Conference Page](https://event.vconferenceonline.com/microsite/html/login.aspx?id=1551&rlp=1)

[ASNR Website] (https://www.asnr.org//annualmeeting)

[Educational Exhibits](https://www.morressier.com/event/asnr2020exhibit/5e5d0be000701ceba4a26941)