Summary
Preserved language function is essential to quality of life. For patients with a brain tumor near putative language
cortex, neurosurgeons may use functional magnetic resonance imaging (fMRI) presurgical language mapping to
assess and mitigate the risks of surgery-induced permanent language deficits. Critical barriers to clinical
deployment of presurgical fMRI are that 1) The validity of conventional fMRI is contingent on the patient’s ability
to perform precisely timed phonological and semantic tasks. But up to 50% of patients assigned to fMRI have
language or other cognitive deficits that affect task performance and may invalidate the mapping. 2) Expertise
in administering language tasks is insufficient in many clinical settings. To tackle these challenges, we propose
a single-group clinical trial to test movie fMRI (in which subjects watch short movie clips while being scanned)
as a novel diagnostic intervention for language mapping in neurosurgical patients. Our primary hypothesis is that
movie-watching relative to task-based and resting-state fMRI can provide comprehensive language mapping in
a greater number of neurosurgical patients, especially those with language deficits. Compared to conventional
fMRI language mapping tasks, movie-watching is predicted to engage more completely the neural networks
supporting language in real life. Movie-watching also improves subject compliance in fMRI scanning. Our
preliminary results demonstrate reduced in-scanner head motion, higher mapping sensitivity in receptive
language areas, and overall higher language mapping specificity, of movie versus conventional task fMRI, in
brain tumor patients undergoing presurgical language mapping (n=34). In this project, Aim 1 will evaluate in
individuals with stroke-induced aphasia (n=80) and healthy controls (n=40), the comprehension and expression
elicited by movie-watching versus conventional paradigms, as a function of language impairment. We predict
that language performance will be comparable, and in most individuals with aphasia superior, for movie-watching
relative to the other paradigms. Aim 2 will assess in neurologically-healthy subjects (n=40), the quality of movie
fMRI language mapping as a function of movie clip, and relative to conventional paradigms, using indices of
language localization sensitivity and specificity, hemispheric lateralization, and proximity to language white-
matter tracts. We predict that for clips inducing the most sensitive and specific synchronized fMRI activation,
movie-fMRI will produce better language mapping quality indices than the other paradigms. Aim 3 will similarly
evaluate movie-fMRI in neurosurgical patients with frontal or temporoparietal gliomas (n=80), as a function of
baseline language function, against the gold standard intra-operative electrocortical stimulation language
mapping, and relative to post-operative language outcome. We predict that movie-fMRI will produce better
language mapping in a large proportion of patients, especially those with language deficits. Study materials will
be publicly released, including optimal movie clips for certain language deficits and fMRI analysis routines.