Research Philosophy
Broadly speaking, I am interested in expressing human cognition and perception
within a mathematical basis. The behavioral neuroscience field contains many
definitions and relationships of these actions. Extending these phenomenological
descriptions, I seek to investigate whether a unique underlying mathematical
formalism based on the developing science of complex systems exists. Such
a functional definition should describe general aspects of both conscious and
subconscious processes and also, hopefully, where the twain meet.
Computational Aspects
A large number and widely distributed spatial scale of neurons are usually in-
volved in cognition, and as such, techniques based from statistical physics may
be useful for its description. On this scale, I am not interested in dynamics at
the cellular level, but rather at the emergent statistical properties based on in-
teractions of its many constituents. It is then hypothesized that cognition can be
viewed as part of this emergence. I propose to use information/probability the-
ory as described by Edwin T. Jaynes and others as the framework for the math-
ematical description. Within this framework, cognition is seen as the combined
probabilities of histories and perception, constrained by intended action/goals.
Experimental Aspects
Due to the emphasis on functional aspects of neural information processing
at multiple time scales, magnetoencephalography (MEG) is used for my neu-
roimaging experiments. It provides a spatial resolution between that of EEG
and fMRI, allowing for discrimination of neural sources and their potentially
differing activity at the millisecond resolution. So far, my MEG related exper-
imental research has focused on sensory processing and resting state no-task
conditions, including the wake-sleep transition. Studies have been conducted
on both normal subjects and patients with mental disorders, in adults and ado-
lescents.
Future Work
My viewpoint of cognition is seen as a time dependent process which is insepara-
ble from sensory aspects/environment and experimental tasks/goals. Therefore,
I propose to develop an appropriate biofeedback MEG experiment, in which
these variables can be accordingly tuned based on current neural states. This
experimental paradigm will hopefully provide a substrate onto which the infor-
mation theoretic model can be based.
Curriculum Vitae
Publications via PubMed
MEG Research
|
