Magnetic resonance imaging combined with diffusion tensor tractography can improve prognosis after ischemic stroke

Abstract

Objective To explore the pattern of functional reorganization in the cortex after corticospinal tract (CST) injury and its relationship with the recovery of upper limb motor function. Methods Fifteen patients with complete paralysis on one side after acute cerebral infarction were studied. Within 1 week after the onset, functional magnetic resonance imaging (fMRI ) and diffusion tensor tractography (DTT) were performed in parallel with timed finger flexion and extension movements in all subjects. The number of nerve fibers in corticospinal tract (CST) in the affected and healthy sides was measured by using Dtv.Ⅱ.R2 software.One and three months later, fMRI was performed while the affected fingers were flexed and extended passively and any cortical activation was observed. In addition, Fugl-Meyer arm motor function scores were assessed one week, one month and three months after the stroke. Results According to the reconstructed nerve fiber number in CST on the affected side, the patients were classified into three types. Type I: the number of newly-built CST nerve fibers is more than 2/3 of that on the healthy side; type II: the ratio is between 1/3 and 2/3; and type III: the ratio is less than 1/3. For typeⅠpatients, blood oxygenation level-dependent fMRI (bold-fMRI) showed initial activation of the bilateral sensorimotor cortex (SMC) and the supplementary motor area (SMA) on the affected side. That was followed by a gradual decrease in the activity in the healthy SMC and an increase in the affected SMC at 1 and 3 months. Among the type II patients bold-fMRI indicated activation of the SMC and SMA on the affected side initially, significant activation of the bilateral SMC and SMA one month later and then stronger activation in the SMC on the healthy side and a weakening of activation in the SMC on the affected side. For type Ⅲ patients, initially the SMA and the posterior parietal cortex were found to be slightly activated. One month later SM1 on the unaffected side was slightly activated, and 3 months later neither the SMC nor the SMA on either side was activated. One week after the onset, the average upper extremity FM scores of the three types of subjects were not significantly different. After one month the three groups′ averages were all significantly different from one another. But after three months the averages for types I and II were again not significantly different, but significantly better than the average of the type III patients. Conclusion Different CST injuries induce different modes of cortical reorganization. The reorganization is a dynamic process, and different activation patterns are closely correlated with clinical prognosis. Key words: Stroke; Functional magnetic resonance imaging; Tractography; Brain reorganization