Background

I studied psychology, inspired by Freud's lectures and Totem and Taboo lying around the local library, at the University of Groningen, and also got my PhD there working with Prof. Ritske de Jong. I wrote my dissertation on cognitive control (the first part of the dissertation was about finding a decent definition for that) and related changes in event-related potentials, and in the time courses of amplitude and phase-locking in specific frequency bands of the EEG.

One interesting result was that when switching between hands, EEG channels over brain areas controlling the hand to which the subject is switching become increasingly phase-locked, in the 20 Hz (beta) range, to other channels over a broad region of the scalp (Gladwin, Lindsen and De Jong, 2006). What the increase means in terms of neural activity remains a question; since the increase in phase locking is associated with a decrease in amplitude at the same frequency, a form of pruning away non-phase-locked activity might be occurring. The shift in phase-locking lateralization only reflected knowing which would be the upcoming hand; it did not covary with reaction time - i.e., transforming the knowledge into effective preparation - as the LRP did for instance.

This rough hypothesis received some support in a subsequent study on motor sequence preparation (Gladwin, 't Hart and De Jong, 2008).

Another phenomenon I found interesting is the transient increase of amplitude during preparation intervals. When switching attention from auditory to visual stimuli, a burst of theta-band (5 - 7 Hz) activity appears over the occipital (visual) region of the brain (Gladwin and De Jong, 2005 ). This second finding inspired an integrate-and-fire neural network model in which selection - via mutual inhibition - was associated with a period of synchronous activity that stopped when the selection process was complete. I liked this little microcosm of science since it linked non-trivial, specific data to an underlying computational mechanism via a mathematical model of a (somewhat) biologically plausible system.

After finishing my Ph.D. work I moved to Utrecht to alleviate a little of my maths deficiency at Utrecht University (which has very good courses and teachers, in particular for the pure maths subjects) and oddjobbed in analyzing corticogram data taken from epilepsy patients before surgery. After some time doing that I was able to participate in fMRI research thanks to Matthijs Vink at the University Medical Center Utrecht. One project I've been involved with looked at the so-called default mode network and what its activity could actually mean, together with setting up correction methods to remove noise from heart rate pulsation, respiration etc. I'm currently also working on genetics analysis at the UMCU.

From July 2007, I started work doing research in a psychiatric clinic in Antwerpen, alongside projects in Utrecht. So far it's been very interesting to actually see how a clinic works, and hear about and encounter patients. The work will mainly involve analyzing changes in the fMRI data of patients who are psychotic when they come to the hospital, but other, behavioral studies are also starting up. So far, my impression is that a certain subset of patients might be well-described as having various kinds of problems with cognitive control; while others have simply responded in a normally adaptive fashion to abnormal situations. I'm hoping that cognitive methods will turn out to be useful for e.g. fine-tuning diagnoses, quantifying clinical impressions, and exploring the causes of negative symptoms and how they result in positive symptoms in schizophrenia.

I currently work at the University of Amsterdam as a post-doctoral researcher in the group of prof. Reinout Wiers. We aim to use various methods to study the neural and cognitive basis of addiction, from the perspective of interactions between automatic and controlled processes. Of particular interest is how various kinds of conditioning not only interfere with, but actually are an element of cognitive control. One method we plan to use is transcranial direct current stimulation, which works to shift the membrane potential, thereby facilitating or inhibiting neural processing.

So the work I seem to be tending towards involves using formal, preferably quantitative models of cognitive / brain function as backbones for psychological and clinical research, together with multiple modalities for probing the causes of behaviour - modalities including both behaviour itself and the various measures of brain activity. Neuroscience research is an example of "cold" computations and calculations meeting "warm" phenomena and experiences. I think this is a central characteristic of neuroscience, and something that makes it especially interesting to me.

Thomas E. Gladwin, Ph.D.

Department of Developmental Psychology
University of Amsterdam

Rudolf Magnus Institute of Neuroscience
Department of Psychiatry
University Medical Center Utrecht

thomasgladwin@hotmail.com