Quantifying the Confidence in fMRI-Based Language Lateralisation Through Laterality Index Deconstruction

Abstract

In epilepsy patients, language lateralisation is an important part of the presurgical diagnostic process. Using task-based fMRI, language lateralisation can be determined by visual inspection of activity patterns or by quantifying the difference in left- and right-hemisphere activity using variations of a basic formula [(L–R)/(L+R)]. However, the values of this laterality index (LI) depend on the choice of activity thresholds and regions of interest. The diagnostic utility of the LI also depends on how its continuous values are translated into categorical decisions about a patient's language lateralisation. Here, we analysed fMRI data from 712 epilepsy patients who performed a verbal fluency task. Each fMRI data set was evaluated by a trained human rater as depicting left-sided, right-sided, or bilateral lateralisation or as being inconclusive. We used data-driven methods to define the activity thresholds and regions of interest used for LI computation and to define a classification scheme that allowed us to translate the LI values into categorical decisions. By deconstructing the LI into measures of laterality (L–R) and strength (L+R), we also modelled the relationship between activation strength and conclusiveness of a data set. In a held-out data set, predictions reached 91% correct when using only conclusive data and 82% when inconclusive data were included. Although only trained on human evaluations of fMRIs, the approach generalised to the prediction of language Wada test results, allowing for significant above-chance accuracies. Compared against different existing methods of LI-computation, our approach improved the identification and exclusion of inconclusive cases and ensured that decisions for the remaining data could be made with consistently high accuracies. We discuss how this approach can support clinicians in assessing fMRI data on a single-case level, deciding whether lateralisation can be determined with sufficient certainty or whether additional information is needed.

Cerebral Polymorphisms for Lateralisation: Modelling the Genetic and Phenotypic Architectures of Multiple Functional Modules

“…A potential problem of all methods of calculating laterality indices, be it from fMRI, fTCD or Wada tests, is that they inevitably lose information on activity in the two separate hemispheres, so that while they provide information equivalent to those shown for motor skill in the present Figure 2b, almost no studies show plots equivalent to Figure 2c. An important exception is the study of Wegrzyn et al [214], who plot L − R against L + R, their Figures 8 and 9 being equivalent to Figure 2c rotated 45 degrees. Unlike Figure 2c there is though no clear divide between the various groups.…”

The relationship between white matter architecture and language lateralisation in the healthy brain

“…Although we employed traditional categorical classifications (left, right, bilateral) to assess language lateralisation, we also employed correlational techniques to investigate the association between LI and white matter characteristics. This approach mitigates the subjectivity concerns associated with the categorical grouping (Wegrzyn et al, 2019; Westerhausen et al, 2006). Even though, our connectometry findings remained consistent across both categorical and correlational analyses, continuous data analysis provided richer information for both BLR/LLD and BLR/RLD groups.…”

Effects of left and right medial temporal lobe resections on hemodynamic correlates of negative and neutral scene processing

“…There was no group difference in anxiety scores (state anxiety: F (2,56) = 3.02, p = .057, η p 2 = 0.10; trait anxiety: F (2,56) = 2.14, p = .127, η p 2 = 0.07), measured using the State‐Trait Anxiety Inventory (Laux et al, 1981 ). The two patient groups did not differ regarding language lateralization ( p = .307, Fisher's exact test) determined using fMRI or Wada test (Wegrzyn et al, 2019 ; Woermann et al, 2003 ), age at epilepsy onset ( t (38) = 0.02, p = .982, d = 0.01), age at surgery ( t (38) = 0.32, p = .749, d = 0.10), time since surgery ( t (38) = 0.39, p = .701, d = 0.13), or postoperative outcome according to Engel class (Engel et al, 1993 ) judged 2 years after surgery ( p = .164, Fisher's exact test). Vision of all participants was normal or corrected to normal.…”

“…STAI‐S: State‐Trait Anxiety Inventory, State subscale; STAI‐T, State‐Trait Anxiety Inventory, Trait subscale (Laux et al, 1981 ). Language lateralization was determined using fMRI or Wada test (Wegrzyn et al, 2019 ; Woermann et al, 2003 ). Outcome of surgery was assessed in follow‐up examination 2 years after surgery.…”

“…The two patient groups did not differ regarding language lateralization (p = .307, Fisher's exact test) (Laux et al, 1981). Language lateralization was determined using fMRI or Wada test (Wegrzyn et al, 2019;Woermann et al, 2003). Outcome of surgery was assessed in follow-up examination 2 years after surgery.…”