From the perspective of neuroscience, written language is an incredible feat. Prompted by reading instruction, the brain constructs specialized circuits to rapidly and automatically translate visual symbols (i.e., letters) into the sounds of language. Indeed, since written language was invented by human societies only a few thousand years ago, it is unlikely that the brain evolved dedicated circuits for written language. Rather, the brain’s capacity to change in response to new experiences - a principle known as “plasticity” - means that a child’s experiences in the classroom sculp the neural circuitry of literacy. While there is consensus that a child’s experiences, such as learning to read, shape brain development, very little is known about the principles governing the relationship between plasticity and learning. Can an enriched learning environment induce changes in the physical connections of the brain? Over what timescale do changes occur: weeks, months or years? Does brain plasticity track learning? Does the brain’s capacity for plasticity and learning diminish as children mature or do circuits remain malleable to experience? We sought to answer these questions by combining an intensive, personalized reading-intervention program for struggling readers with longitudinal MRI measurements of brain connections (i.e., white matter).

33 struggling readers between 7 and 12 years of age (most with a dyslexia diagnosis) were recruited for a summer reading intervention program that involved one-on-one instruction in the building blocks of literacy. A unique aspect of this study was the intensity of the intervention: four hours a day, five days a week for eight weeks. Behavioral measures of children’s reading abilities demonstrated steady growth in all aspects of reading: accuracy, speed, fluency and comprehension. On average, children improved by about one grade level, amounting to a substantial and meaningful gain over eight weeks of summer. 

MRI measures of the white matter – the brain tissue containing connections between distant brain regions – revealed a surprising capacity for plasticity occurring over a rapid timescale. Changes in white matter structure emerged within the first weeks of the intervention program and continued throughout the intervention period. These changes in brain connectivity were found to track the learning process and occurred in synchrony with improvements in reading skill. Beyond the core connections of the brain’s reading circuitry, plasticity was also observed across a widespread network of tracts, indicating far more extensive changes than were originally anticipated.

These findings have far-reaching implications for education, particularly, for children with dyslexia. First of all, teachers play a critical role in shaping brain development. Second, neurobiological differences in learning disabilities like dyslexia should not be presumed static traits. When placed in the right learning environment, a child’s brain connections and reading skills can both change substantially over a relatively short period of time. Finally, we observed substantial reading growth and white matter plasticity across the entire age range (7 – 12 years) indicating that older children can still benefit from targeted educational interventions.

The discoveries from this research were grounded in methodological innovations including new software algorithms for (1) processing MRI data, (2) extracting quantitative measures of white matter plasticity, (3) visualizing brain connections and (4) storing and sharing results. All the software we developed for this project has been released as open-source software through the lab GitHub page: https://github.com/YeatmanLab. Notable open-source software packages that emerged from this work were: (1) Updates to the Automated Fiber Quantification software to accommodate longitudinal data and novel modeling approaches (https://github.com/YeatmanLab/AFQ); (2) A Python version of the popular Automated Fiber Quantification software package supporting integration with cloud-computing platforms (https://github.com/yeatmanlab/pyAFQ); (3) A new browser-based platform for analysis, visualization and sharing of diffusion MRI data (https://github.com/YeatmanLab/AFQ-Browser).

Last Modified: 05/30/2020
Modified by: Jason Yeatman