CURRENT RESEARCH
Field potentials recorded from the brain (invasive or non-invasively) are non-stationary signals that represent the dynamic interactions between local neural populations. Synchronized interactions between populations then manifest as oscillatory bursts in the field potentials. These bursts are called neuromodulations and have been recognized by clinicians as bursts of rhythmic activity that wax and wane when eyeballing the signal. Further, they are temporally sparse and transient. On the other hand, unsynchronized interactions contribute to background activity that is featureless and unstructured.
Neuromodulatory patterns command much interest because of their role as potential biological markers in the early detection of neurological diseases and as neural patterns carrying cognitive content. While classical signal processing techniques attempt to identify these oscillatory bursts using spectral representations of the signal, they are limited by the inherent time-frequency trade-off. Moreover, such analyses demand more specialized methods that incorporate the neurophysiology of signal generation and high-time resolution. In this project, we aim to address these issues to classify and characterize neuromodulations from field potentials. We do so in an unsupervised manner using information theoretic measures, sparse-coding techniques and point process modeling.
Working memory is a complex cognitive process involved in encoding, storing and retrieving sensory information. The process can be divided into 3 sub-phases: memory creation, maintenance and retrieval. Neurons in the prefrontal cortex (PFC) have been shown to maintain transiently elevated activity during delay period in cued choice memory guided tasks. However, another school of thought supports, additionally, the involvement of population activity that are exemplified in the complex dynamics underlying the field potentials. These dynamics correspond to oscillatory bursts (i.e., neuromodulations) in the gamma and beta frequency bands.
Through this study, we aim to summarize the involvement of oscillatory dynamics as the brain transitions between the different WM states during task execution. We sought to validate our hypothesis that neuromodulatory events indicate the emergence of task variables, are representative of the stimulus, as well as encode correlates of executive control; and therefore, they may be predictive of behavior. Simultaneously, we asked two critical questions: 1) what is the nature of encoding of neuromodulations? and 2) how important is time-resolution for abstracting relevant information from field potentials?
Sleep spindles result from interactions between the thalamic and cortical neurons during the NREM2 stage. Studies show that these waxing and waning episodes of field potentials may have an implied role in memory consolidation, cellular plasticity and neuronal development besides serving as important markers in several neuronal pathologies. For these reasons, accurate spindle scoring of polysomnographic signals is important and has garnered interest in automating the tedious process of scoring via visual inspection. In this project, we employ the transient model for neuromodulatory events as a generalized automatic sleep spindle detector.
FUNCTIONAL CONNECTIVITY ANALYSIS
Will be updated soon..