Diffusion tensor imaging can predict long-term functional outcomes after ischemic stroke

Abstract

Objective To quantify and predict long-term motor outcomes after ischemic stroke using diffusion tensor imaging (DTI). Methods Ten patients with middle cerebral artery infarction were prospectively studied using DTI within 3 months and 1 year after the onset. A region-of-interest-based analysis was performed for the fractional anisotropy (FA) of the pyramidal tract in the cerebral peduncles (CP) and the pontocerebellar tract in the middle cerebellar peduncles (MCP). Neurological function was evaluated using the National Institutes of Health′s stroke scale (NIHSS) and the degree of paresis was assessed at the same time using paresis grading (PG). During the last follow-up, the functional outcome, ability in the activities of daily living and balance function were evaluated using the modified Rankin scale (mRS), functional independence measures (FIMs) and the Brunel balance assessment (BBA) respectively. Results The average fractional anisotropy on the affected side of the CP was significantly lower than that of healthy persons at both 3 months and 1 year after onset. The FA of the MCP was significantly higher than on the healthy side. The ratio of the FA on the affected side to that on the unaffected side (rFA) in both the CP and MCP was significantly correlated with the paresis grading within 3 months, and significantly correlated to the NIHSS score, the upper extremity′s PG, total PG, mRS score and FIM score at the end of the follow-up. At the end of follow-up, the rFA of the MCP correlated in a better tendency with the lower limb PGs and balance ability than that of the MCP, although the differences were not significant. Long-term motion function was not significantly correlated with a patient′s age, motion dysfunction, size of focus, or the rFA of the CP and MCP. However, lower limb function was significantly correlated with the size of the focus and the rFA of the MCP, while upper limb mobility was significantly correlated with age, the size of the focus and the rFA of both the CP and the MCP. The area under the receiver operating characteristic (ROC) curve for the rFA in the MCP was greater than the area for the rFA in the CP. The cutoff point for the rFA in the MCP for a good lower extremity motor outcome was determined to be 0.826 (sensitivity of 71%, specificity of 100%). For the CP the cutoff was 0.77 (sensitivity of 57%, specificity of 100%). Conclusions DTI can predict the lower extremity motor outcomes and prognosis after stroke. Key words: Ischemic stroke; Diffusion tensor imaging; Middle cerebellar peduncles; Motor recovery; Fractional anisotropy