Motor deficits of the upper limb (UL) are one of the most common consequence after stroke and affect up to two third of stroke survivors. Post-stroke disturbances in the balance of excitatory, inhibitory, and frequency-specific mechanisms in the motor network lead to motor deficits and recovery is aimed at re-setting the imbalance to normal (i.e. brain plasiticity). As there is a post-stroke "sensitive period" (within 6 months poststroke) that most recovery from impairment would occur, accurate prediction of rehabilitation outcome at an early stage after stroke becomes crucial for physicians to facilitate the proper treatment to maximize the gain from therapy by selecting individualized treatment. Therefore, in this study, we presented a machinery for predicting the recovery at administration by using Dynamical Causal Modelling and machine learning. We hypothesized that the poststroke patterns of frequency-specific excitatory and inhibitory connectivity in the motor network may reflect the ability to recovery and could serve as good neuromarkers for predicting the recovery. We extracted the post-stroke electroohysiological features from EEG data, which were recorded when patients were moving their affected hand. These network features then went into the classifiers for two-class classification: good and general improvement. We will discuss the features and demographic factors that have an impact on the prediction accuracy and compare the performance of this prediction machinery with others to show its merits. We expect that the results of this study will lead to a better outcome for patients to gain from the rehabilitation.
|Effective start/end date||1/08/16 → 31/07/17|
In 2015, UN member states agreed to 17 global Sustainable Development Goals (SDGs) to end poverty, protect the planet and ensure prosperity for all. This project contributes towards the following SDG(s):