Functional near-infrared spectroscopy is in moderate accordance with functional MRI in determining lateralisation of frontal language areas

Abstract

Purpose Understanding language dominance is crucial in pre-surgical evaluation of patients with epilepsy and in patients having a tumour close to the language area. Functional magnetic resonance imaging (fMRI) studies are well established in evaluating language dominance. Functional near-infrared spectroscopy (fNIRS) is a non-invasive optical imaging modality that offers a convenient and affordable technique to image language-related cortical areas. This study investigates the agreement between results from task-based fMRI and fNIRS in determining language lateralisation. Methods Language laterality indices LIs were calculated from both fMRI and fNIRS measurements of the same individual volunteers by using an identical paradigm. Statistical measures of percentage agreement and kappa value have been calculated for testing agreement and reliability. Results A correlation analysis of the LI values shows a good correlation with r = 0.677 at p < 0.05. Statistical comparison of both fMRI and fNIRS methods for language lateralisation yielded a percentage agreement of 90% and a moderate kappa value of κ = 0.621. Conclusion Our study suggests that fNIRS is in moderate accordance with fMRI in determining lateralisation of the frontal language areas. It implies that the optical imaging technique can provide additional information on functional lateralisation of frontal language areas.

Keywords: fMRI; fNIRS; language lateralisation; laterality index.

Figure 1.

Inferior frontal gyrus montage used for the study. S1–S8 represent the sources and D1–D8 represent the detectors.

Figure 2.

(a) and (b) Three-dimensional-rendered image of haemodynamic activity measured by functional near-infrared spectroscopy on the right and left hemisphere, respectively. (c) and (d) Activity mapped by functional magnetic resonance imaging on the right and left hemisphere, respectively. SPM: scanning probe microscope; HbOxy: oxyhaemoglobin.

Figure 3.

The average oxyhaemoglobin (HbOxy) signal showing an increase during the task (Condition 2) and a decrease as the 20-second task is being completed by the participant.

Figure 4.

Linear correlation between the LI values of functional magnetic resonance imaging (fMRI) and functional near-infrared spectroscopy (fNIRS). An r value of 0.67 indicates good correlation between the two methods of calculation of LI.

Figure 5.

The top graph shows the difference in oxyhaemoglobin signal obtained from left and right hemispheres averaged across the respective channels. The bottom graph shows the difference in blood oxygen level-dependent time series averaged across the regions of interest from both hemispheres.