Advances in Cognitive and Neural Studies: Bridging Neural Mechanisms and Cognitive Processes

Abstract

The explanation of how neural activity results in human cognition needs reasons of analysis that can decode the wide spectrum dynamics of the brain spatiotemporal dynamics. This paper is a review of the recent developments in the fields of cognitive and neural research, and specifically discusses neural signal processing algorithms that span the gap connecting lower-level neural processes, and higher-level cognition. The current methods, such as Independent Component Analysis, time-frequency decomposition based on Wavelet Transform, and graph-theoretic models of connectivity allow a perfect extraction of neural features that are related to perception, memory, attention, learning, and decision-making. The classification and prediction of cognitive states regarding EEG, MEG, and fMRI signals are also more advanced through machine learning and deep-neural architecture, including the SVMs, CNNs and hybrid CNN–LSTM models. Other levels of interpolation to neurons are available via integrative computational models such as spiking neural model and Bayesian cognitive model. Combined with additional algorithmic advances, these help to enable more precise, interpretive and biologically plausible indices between brain activity and cognition. This work, by bringing together technologies available in neuroscience, cognitive science, artificial intelligence, and biomedical engineering, highlights the additional opportunities of neurocomputational methods to deepen the relationship between the minds and guide the neurotechnology of the future.