Koichi Sameshima.

According to the WHO, more than one billion people in the world experience disability, which is not only an issue of public health, but also of human rights. In 2015, the Global Burden of Diseases showed worldwide epidemiological evidence that 74% of total years lived with disability are related to conditions for which rehabilitation could be beneficial. For this reason, the development of new therapies and rehabilitation research are one of the main objectives of the WHO Global Action Plan for 2014-2021. However, the lack of knowledge of the biological mechanisms involved in the functional rehabilitation process is one of the main limitations faced in the construction of a solid knowledge. There is evidence that brain plasticity is the central mechanism involved in the recovery process of patients with deficits of different etiologies. Although much progress was made in understanding brain plasticity, little was incorporated into rehabilitation practice, mainly due to the scarcity of study designs that allowed the translation of knowledge. Therefore, we propose a study with four groups of subjects (with stroke, SCI, amputations and osteoarthritis) to understand the neuroplasticity mechanisms involved in motor rehabilitation, using TMS, fNIRS and high density EEG before and after the rehabilitation treatment. With this project, we expect to better understand the mechanisms of neuroplasticity involved in the rehabilitation process, and to develop "transdiagnostic" neurophysiological biomarkers, which are of great relevance for the scientific and therapeutic improvement of rehabilitation.

Clinical and basic neuroscientists using functional magnetic resonance imaging, electroencephalography or magnetoencephalography are all in need of straightforward tools to determine causal relationships among brain structures. Over the past decade, functional neuroimaging or brain mapping has enjoyed enormous success in identifying functionally specialized areas in the brain both in systems and cognitive neuroscience. This approach, however, relies solely on functional mapping and is hence incapable to expose more fundamental principles that underlie brain functional architectures. 
We employ partial directed coherence (PDC), a more neurophysiologically meaningful measure of the neural interactions among brain structures, which is based on the Granger causality concept. This concept not only allows results comparison using the data collected using different technologies but also is a strong candidate for replacing the currently popular pairwise correlation techniques in the analysis of neural dynamics because it increases both the sensitivity and the specificity of effective connectivity diagnosis. As such it is able to reduce multivariate data analysis misinterpretations, either for BOLD and for electrophysiological signals, because it takes into account a larger number simultaneous measured neural structures (variables). More importantly, it is not only able to disclose directed interactions in neural systems, but is also able to detect the presence of feedback. Because of the simultaneous consideration of many structures, our project contemplates further analysis and visualization by means of graph theoretical and computer based visualization methods to facilitate the identification of information flow among brain structures.