Changes in Clinical and Network Functional Connectivity Parameters in Motor Networks and Cerebellum Based on Resting-State Functional Magnetic Resonance Imaging Data in Patients with Post-Stroke Hemiparesis Receiving Interactive Brain Stimulation Neurotherapy

Abstract

Introduction. Interactive brain stimulation (IBS) neurotherapy is an advanced neurofeedback technology (NFB) that involves the organization of a feedback “target” based on signals recorded by functional magnetic resonance imaging (fMRI) and electroencephalography (EEG). The NFB allows patients to volitionally self-regulate their current brain activity and may therefore be a useful treatment option for diseases with altered activation and functional connectivity (FC) patterns. Our objective was to assess the effects of IBS on the FC changes in motor networks and correlations between clinical and network parameters in patients with post-stroke hand paresis. Materials and methods. Patients with a history of stroke 2 months were randomized into a main group (n = 7) and a control group (n = 7). All the patients followed the stroke physical rehabilitation for 3 weeks. The main group received IBS training, where the patients learned to imagine movements of the paretic hand trying to amplify the fMRI signal from the primary motor cortex (M1) and the supplementary motor area (SMA) on the lesion side with simultaneous desynchronizing the μ- and β-2 EEG rhythms in the central leads. Clinical tests and MRI were performed prior to and immediately after the treatment. FC matrices were constructed using CONN software based on resting-state fMRI data. Results. By the end of the training, M1–M1 functional connectivity in the control group weakened, while no changes were observed in the main group. The FC strength was positively correlated with the grip strength (ρ = 0.69; p 0.01) and with the results of the Box and Blocks test (BBT score, ρ = 0.72; p 0.01) and the Fugl-Meyer assessment for upper extremity (FM-UE score, ρ = 0.87; p 0.005). Ipsilesional SMA connectivity with contralesional cerebellum weakened (p 0.05 in the main group). Its strength was negatively correlated with the BBT and FM-UE scores (both tests ρ = –0.44; p 0.05). Conclusions. Volitional control of M1 and SMA activity in the lesion hemisphere during the post-stroke IBS training alters the architecture of the entire motor network affecting clinically significant FC types. We studied a possible mechanism of this technology and its potential use in treatment programs.