David E. Warren, PhD

Hubs in brain networks

Most networks are configured hierarchically: hub regions have greater importance in organizing or facilitating network activity than other regions. This principle of hub-centered construction has been applied to the resting-state networks of the brain, but there is little consensus in the literature about what regions are true hubs due to the variety of methods and measures used in their identification. In a collaborative investigation with Dr. Steve Petersen's lab at Wash U, we tested a novel theory of brain network structure by using neuropsychological methodology to evaluate potential hubs in the brain. Brain-injured patients with damage to two types of potential hub locations were evaluated with the hypothesis that damage to brain hubs should greatly impair cognition. Critically, patients with damage to hubs that were connected to many distinct brain systems showed gross cognitive impairment that was disproportionate to lesion size or location, while patients with damage to other potential hubs showed little cognitive change.

vmPFC and false memory

The ventromedial prefrontal cortex (vmPFC) has long been associated with decision-making in complex contexts. However, recent work has suggested that the vmPFC may be involved in schematic memory operations. I investigated this claim by testing patients with damage to the vmPFC using the Deese-Roediger-McDermott (DRM) paradigm. Intriguingly, vmPFC patients showed reduced false memory, suggesting that the vmPFC contributes to false memory in healthy individuals by filling in contextually appropriate information even if it was not studied. In ongoing work, we are testing the influence of vmPFC on other memory phenomena including false memory and schematic memory.

Relational memory and hippocampus

Relational memory describes the ability to bind together arbitrarily related bits of information rapidly and automatically, and we believe that this ability underlies our rich episodic memories among others. Studies of patients with severe amnesia have shown that injury to a relatively small brain structure, the hippocampus, is sufficient to severely impair relational memory while leaving other types of memory relatively intact. We continue to investigate the implications of hippocampal damage and dysfunction in new tests of memory and cognition.

MTL and on-line processing

The medial temporal lobes (MTL) play a key role in forming new memories. However, recent research shows that loss of MTL and/or hippocampus can profoundly affect ongoing cognitive processes. My work has tied hippocampal damage to deficits in on-line comparison of visual stimuli, an unexpected finding given that the hippocampus is most often associated with forming enduring relational memories. I am continuing to explore the nature of cognitive deficits related to hippocampal processing.

Word learning in amnesia

Anterograde amnesia is a generic impairment of learning factual, episodic, or what has been referred to as relational information. However, it has recently been suggested that certain types of training might help severely amnesic patients to learn new words. If true, this would be a very important advance for memory researchers and for clinicians. I developed a test of this new method with severely amnesic patients. Unfortunately, we were unable to replicate an earlier finding supporting this claim. Instead, my research indicates that amnesia does generally impair new learning of relations between, for example, words and images, regardless of the teaching technique used.

Age and eye movements

Experience of our visual environment relies first on our eyes, and the ability to move our eyes is known to change with age. I am currently investigating changes in eye movements due to healthy aging, and additional changes due to dementia of the Alzheimer's type.