Functional Near-Infrared Spectroscopy (fNIRS) is a characteristically functional neuroimaging technique which enables us to measure the daily tasks related to cortical activation including gait and postural task. Using fNIRS, it was found that the medial sensorimotor and supplementary motor area play an important role in gait and postural control in healthy subjects. In addition, it was also revealed that the individual balance ability was correlated with the cortical activation in the supplementary motor area during the postural task. These findings supported the notion that the supplementary motor area is one of the key structures for balance recovery in stroke patients. Not only can fNIRS effectively monitor the functional reorganization of the central nervous system, but fNIRS has also been used as a therapeutic tool. With recent advances in technique enabling real-time decoding of brain activity, functional neuroimaging can now be used as a neurofeedback tool, in which the voluntary modulation of cortical activation is available. After we developed a working fNIRS mediated neurofeedback system and confirmed its neuromodulation effect in healthy subjects, we investigated its clinical efficacy as a therapeutic tool for augmenting the functional recovery after stroke. Our pilot randomized control study revealed the promising result that neurofeedback intervention could improve finger function in chronic stroke patients including patients with moderate to severe paresis. These findings provide a new therapeutic possibility for those patients who gain only limited functional recovery from conventional rehabilitative interventions in the chronic stage.

The purpose of this study is to evaluate MRI findings for the shoulder pain in hemiplegic stroke patients in relation to clinical characteristics. We studied 18 hemiplegic patients with first-ever stroke presenting with shoulder pain in the affected side (mean age±SD=67.6±10.1years ; 8 men and 10 women ; 12 right and 6 left hemiplegia). All patients had shoulder pain during passive movements and 4 also had pain at rest. The mean duration from stroke onset to MRI was 67±42 days. MRI revealed abnormal findings in all patients. Tendinosis of the long head of the biceps and supraspinatus tendon injuries were most frequently found. The tendinosis of the long head of the biceps was related to hemihypesthesia and a reduced range of motion for external rotation of the shoulder. The supraspinatus tendon injuries were related to older age and lower Fugl-Meyer (FM) and Functional Independence Measure (FIM) scores. The mean number of abnormal findings per patient was 3.2±1.4. The patients with more than 3 abnormal findings were significantly older, had shorter duration from stroke onset and lower FM and FIM scores than those with less findings. It was suggested that those patients with more severe paresis might have more abnormal findings on their MRIs for shoulder pain.