Title: Plataforma de computação neurocognitiva:protótipo na reabilitação do zumbido através da regulação da emoção concomitante com a estimulação transcraniana por corrente diret
Abstract: Brain cognitive networks consolidate emergently according to but not limited to individual lifelong experiences related to emotion, stress, fear, attention, learning, society, language, religion, culture, and family environment.Therefore, cognitive functioning, causality versus contingency issues, brain networks, intervention protocol, and rehabilitation effectiveness are challenges in cognitive and neuroscience sciences.Additionally, data-driven knowledge supports discovering patterns, creating decision-making strategies, and making predictive and perspective analytics possible.However, cognitive rehabilitation is also desired by updating knowledge and autonomous learning through new conditions due to individualized emergent brain consolidation.Eventually, digital infrastructure and computing performance are essential to make the developed models accessible, affordable, and updatable for the healthcare system.One of the practical solutions that can collectively address all challenges is a cognitive computing platform.Cognitive Computing(CC) is an emerging knowledge niche grounded on cognitive science, neuroscience, data science, and computer technologies.Architecturally, I propose a Cognitive Computing platform constructed from micro-module networks called neurocognitive computing.Neurocognitive Computing (NcC) is a self-regulatable cognitive computing unit that can learn and be trained in real-time to generate cognitive system mechanisms individually or in a complex network.In this 8-year project, I tried to make an integrated Cognitive Rehabilitation Platform (iCRP) as neurocognitive computing to navigate cognitive impairment (emotion dysregulation) and rehabilitation with transcranial electrical stimulation techniques for tinnitus patients.In iCRP, we conceptualized and partially developed neurocognitive computing segments in parallel, from theoretical cognitive modeling, clinical trial, and un/supervised data-driven mining to real-time deep learning.I tried to document the cognitive science section of the project concisely and profoundly, developing theoretical and experimental toolboxes for unsupervised pattern recognition and exploring the data-driven methodologies with metric performance evaluations to fuel deep reinforcement learning in the future.