Brain Function and Upper Limb Outcome in Stroke: A Cross-Sectional fMRI Study

Abstract

Objective: The nature of changes in brain activation related to good recovery of arm function after stroke is still unclear. While the notion that this is a reflection of neuronal plasticity has gained much support, confounding by compensatory strategies cannot be ruled out. We address this issue by comparing brain activity in recovered patients 6 months after stroke with healthy controls.

Methods: We included 20 patients with upper limb paresis due to ischemic stroke and 15 controls. We measured brain activation during a finger flexion-extension task with functional MRI, and the relationship between brain activation and hand function. Patients exhibited various levels of recovery, but all were able to perform the task.

Results: Comparison between patients and controls with voxel-wise whole-brain analysis failed to reveal significant differences in brain activation. Equally, a region of interest analysis constrained to the motor network to optimize statistical power, failed to yield any differences. Finally, no significant relationship between brain activation and hand function was found in patients. Patients and controls performed scanner task equally well.

Conclusion: Brain activation and behavioral performance during finger flexion-extensions in (moderately) well recovered patients seems normal. The absence of significant differences in brain activity even in patients with a residual impairment may suggest that infarcts do not necessarily induce reorganization of motor function. While brain activity could be abnormal with higher task demands, this may also introduce performance confounds. It is thus still uncertain to what extent capacity for true neuronal repair after stroke exists.

Fig 1. Surface reconstruction of a single subject with the anatomical motor segments depicted by different colors.

Fig 2. Axial structural T1-weighted MRI scans at the level of maximum infarct volume for each patient performed at the time of the fMRI session.

Fig 3. Histogram of incidence of isometric contractions (MM) in the contralateral hand defined by a score consisting of the correlation of the electromyography (EMG) signal measured at the extensor muscles of the contralateral hand during the task with the task boxcar multiplied by the average % of maximal voluntary contraction (%MVE¯) (measured before the session in the scanner) of the muscle during the session.

Patient-scores are depicted in black. Control-scores are depicted in grey.

Fig 4. Mean results for Amplitude and Force tasks for the unaffected and affected hand for patients and controls.

Bars show the mean beta per ROI (±1 SD) cerebellum, PM, SMA, postcentral gyrus, precentral gyrus and insula for the left (affected) and right (unaffected) hemisphere (LH, RH). Patients’ T-maps were flipped so affected hand was always the right hand.

Fig 5. Mean results for Laterality Index for Amplitude and Force tasks for the unaffected and affected hand for patients and controls.

Mean LI per ROI (±1 SD) cerebellum, PM, SMA, postcentral gyrus, precentral gyrus and insula.

Fig 6. Group activation move vs. rest for 4 sessions, unaffected amplitude (UA), affected amplitude (AA), unaffected force (UF), and affected force (AF) between patients (P) and controls(C).

For illustration purposes threshold was set at T>5 uncorrected.