Description of Lab Projects in Progress
The overall goal of my research is to understand mechanisms of brain plasticity and development using functional magnetic resonance imaging (fMRI). In addition to examining typical populations, my work also investigates atypical conditions such as dyslexia (reading disorder) and attention deficit hyperactivity disorder (ADHD). It, therefore, has direct implications for the diagnosis and treatment of developmental disorders.
A major focus of my research has been on learning and age-related changes in the orthographic (spelling), phonological (rhyming) and semantic (meaning) representations that support reading and other linguistic processes. Our work is unique because we use several baselines to control for non-linguistic processing, parametric manipulations of difficulty, and task comparisons, to determine if brain activation is specific to lexical processing. We also combine cross-sectional with longitudinal designs so that we can more effectively examine developmental trajectories using non-linear growth curves. Finally, we dissociate effects of development from effects of skill by using a multiple regression approach.
This multifaceted approach to research has supported several general conclusions about the brain development of orthographic, phonological and semantic processing. First, these systems appear to become more elaborated with development. Age-related increases in activation are seen in brain regions implicated in processing each of these types of representation. Interestingly, however, the effects seem to be more complicated for the fusiform gyrus where there are decreases in activation of posterior regions (implicated in letter processing) coupled with increases in anterior areas (implicated in word processing). This pattern suggests a shift in focus from smaller to larger orthographic units in reading development. Second, the mapping between the orthographic, phonological and semantic systems seems to become more efficient as there are developmental increases in activation in the brain regions involved in their integration. We have thus far identified two such integration regions in the inferior parietal lobule, one that integrates orthographic with phonological information and the other that integrates orthographic with semantic information. Third, the orthographic, phonological and semantic systems appear to become more specialized because there is less overlap in the systems for processing auditory and visual information with development. There is also greater left hemisphere lateralization as shown by increases in activation in left inferior frontal gyrus. We have used effective connectivity to show that the inferior frontal gyrus exhibits age-related increases in top-down modulation on posterior brain regions involved in integrating task relevant information. Fourth, compensatory mechanisms appear to be used less with development as there are age-related decreases in brain regions that are not critical for performing a specific task. For example, there is less activation of semantic representations in rhyming or spelling tasks in adults compared to children.
An increasing focus of my research involves an examination of how the brains of dyslexic children are different from normal populations. In particular, I have a particular interest in the degree of brain plasticity in these groups. We have shown that dyslexic children exhibit less activation than controls in several nodes of the language network, but also that there are interesting differences in age-related and performance-related effects between the groups. Dyslexics who have higher accuracy levels show greater involvement of the right hemisphere, suggesting that it is involved as a compensatory mechanism. There are also age-related decreases in critical nodes in the left hemisphere suggesting that dyslexia is a progressive disorder. It is well known that dyslexia is a heterogeneous population, so in the future we will examine different subtypes of dyslexics, e.g. those with single word reading deficits versus those with more generalized language deficits, to determine if they have different neural ‘ signatures’. We will also examine whether these groups represent a developmental delay versus deviance by comparing them to younger controls matched for reading skill.
One of the central questions in developmental science is whether aspects of development are different from learning. My research is increasingly using learning designs because we can directly examine brain plasticity and how brain plasticity is different in younger versus older children. For example, we have recently examined neural changes in English speakers when learning the meaning of Chinese characters. Transfer of semantic knowledge was also measured by examining the processing of novel characters that shared a semantic radical (part of a character that gives a clue to word meaning, e.g. water for lake) with trained characters. We found that there were learning related increases in activation in several brain regions implicated in English word processing, but there were also increases in regions implicated in visual-spatial processing. In order to have more complete control over the experimental stimuli in our learning experiments, we also developed an artificial orthography to examine reading acquisition. Our results bear heavily on ongoing controversy regarding whether to emphasize ‘explicit’ instruction of the alphabetic principle (mapping between letters and sounds) or ‘implicit’ instruction involving whole words. Adults taught to read our artificial orthography with implicit instruction on whole words relied on declarative memory processes and acquired greater knowledge of large orthographic units (visual-patterns) with training, whereas those taught with explicit instruction on the alphabetic principle revealed greater proceduralization of smaller orthographic units (letters).
Interestingly, both types of learning result in increases in activation in the left inferior frontal gyrus. In the future, we will compare these learning processes in both typically developing and dyslexic children. We will test a model of reading acquisition that posits a developmental shift from sensitivity to mappings between larger units (e.g. whole words) in children, to sensitivity to smaller units (e.g. letters) in adults. This model predicts that children will be more sensitive to visual-pattern information in larger orthographic units, whereas adults will be more sensitive to alphabetic information spanning smaller orthographic units. An extension of this model to reading disorders posits that dyslexics will have pronounced deficits in acquiring mappings for the smaller orthographic units because this relies on a more advanced process.
Another area of my research involves developmental changes in executive functioning (e.g. response inhibition) and spatial selective attention (e.g. visual search). We have shown that there are smaller age-related differences for visual search than for response inhibition, and have suggested that this could be due to slower maturation of the frontal lobes critical to executive functioning. We have also shown that ADHD children exhibit larger brain abnormalities during executive functioning than spatial selective attention, suggesting that a central deficit in ADHD may be in response inhibition. More generally then, it may be that later developing cognitive abilities are more susceptible to developmental disorders. Because our previous research has shown pronounced deficits in executive functioning in ADHD children as measured by response inhibition, we will examine whether there is a more general neural deficit in executive functioning that also involves spatial working memory (e.g. 2-back task). As with dyslexia, ADHD is a heterogeneous population and there are probably multiple pathways to the disorder. In fact, other investigators have shown that some ADHD patients have an executive functioning deficit whereas others have a reinforcement circuitry deficit, so we will examine whether networks in the prefrontal cortex and striatum are differentially effected in these populations by manipulating the delay-of-reinforcement gradient in executive functioning tasks. Other investigators have also shown that methylphenidate, a psychostimulant medication for ADHD treatment, seems to have greater influence on dopamine in brain regions involved in executive functioning, so we will examine whether those with an executive functioning deficit will show greater normalization in this network with treatment.
