Early functional magnetic resonance imaging activations predict language outcome after stroke

Abstract

An accurate prediction of system-specific recovery after stroke is essential to provide rehabilitation therapy based on the individual needs. We explored the usefulness of functional magnetic resonance imaging scans from an auditory language comprehension experiment to predict individual language recovery in 21 aphasic stroke patients. Subjects with an at least moderate language impairment received extensive language testing 2 weeks and 6 months after left-hemispheric stroke. A multivariate machine learning technique was used to predict language outcome 6 months after stroke. In addition, we aimed to predict the degree of language improvement over 6 months. 76% of patients were correctly separated into those with good and bad language performance 6 months after stroke when based on functional magnetic resonance imaging data from language relevant areas. Accuracy further improved (86% correct assignments) when age and language score were entered alongside functional magnetic resonance imaging data into the fully automatic classifier. A similar accuracy was reached when predicting the degree of language improvement based on imaging, age and language performance. No prediction better than chance level was achieved when exploring the usefulness of diffusion weighted imaging as well as functional magnetic resonance imaging acquired two days after stroke. This study demonstrates the high potential of current machine learning techniques to predict system-specific clinical outcome even for a disease as heterogeneous as stroke. Best prediction of language recovery is achieved when the brain activation potential after system-specific stimulation is assessed in the second week post stroke. More intensive early rehabilitation could be provided for those with a predicted poor recovery and the extension to other systems, for example, motor and attention seems feasible.

Neurophysiological Markers of Language Recovery in Subacute Stroke

“…Bilateral SFG activity at the subacute stage was also related to aphasia recovery, consistent with findings of increased subacute activity in frontal regions in previous longitudinal studies on aphasia recovery (Fernandez et al, 2004;Saur et al, 2006;Saur et al, 2010). However, in the present study, we observed a negative correlation between left SFG subacute activity for pseudowords and single word comprehension at the chronic stage.…”

“…Nevertheless, the degree of left insula/IFG activity in an individual was related to overall recovery capacity. This finding is in general agreement with the observation that left frontal activity in the subacute stage is associated with improved word comprehension (Heiss et al, 1999) and predictive of global language improvement (Saur et al, 2010).…”

“…Despite these limitations, a follow-up study including an additional seven patients (total 21) who performed a different but related sentence comprehension task showed striking results (Saur et al, 2010). Overall, 86% of stroke patients with aphasia at 2 weeks were correctly predicted as having good or poor language recovery at over 6 months based on fMRI data in addition to language performance and age (with similar results for predicting improvement).…”

“…Improvement scores from the language assessments were calculated using the chronic scores minus the subacute score and dividing by the amount of possible maximal recovery (maximum test score -subacute score) or maximal impairment (subacute score -minimum test score) if the chronic score is lower than subacute. These percentage maximal recovery or impairment (PMRI) values were used in the correlational analyses and this approach is based on similar methods that have been used previously to measure language improvement in aphasia (see Lazar, Speizer, Festa, Krakauer, & Marshall, 2008;Saur et al, 2006;Saur et al, 2010) Separate whole brain analyses were conducted using a factorial ANOVA with the group as the between-subject and conditions as the within-subject factor. Four analyses were carried out, using a 2 (subacute, chronic) x 2 (word, pseudoword), a 2 (subacute, chronic) x 3 (concrete, abstract, pseudoword), a 3 (subacute, chronic, control) x 2 (word, pseudoword) and a 3 (subacute, chronic, control) x 3 (concrete, abstract, pseudoword) factorial.…”

Multivariate Classification of Blood Oxygen Level-Dependent fMRI Data with Diagnostic Intention: A Clinical Perspective

“…For example, Du et al 55 combined both task-and task-free fMRI in schizophrenia in a small study. Additionally, combinations of fMRI measures with volumetric data, 41,[48][49][50]63,76,[78][79][80][81]86,89 DTI, 46,49,92 as well as genetics 42 and behavioral data, 40,41,50,76 have been used as features in MVPA analyses. However, results reported so far do not allow verified statements about the benefit of such multimodal acquisitions.…”