Login / Signup

Associative white matter tracts selectively predict sensorimotor learning.

Sophia Vinci-BooherDaniel J McDonaldE BerquistFranco Pestilli
Published in: Communications biology (2024)
Human learning varies greatly among individuals and is related to the microstructure of major white matter tracts in several learning domains, yet the impact of the existing microstructure of white matter tracts on future learning outcomes remains unclear. We employed a machine-learning model selection framework to evaluate whether existing microstructure might predict individual differences in learning a sensorimotor task, and further, if the mapping between tract microstructure and learning was selective for learning outcomes. We used diffusion tractography to measure the mean fractional anisotropy (FA) of white matter tracts in 60 adult participants who then practiced drawing a set of 40 unfamiliar symbols repeatedly using a digital writing tablet. We measured drawing learning as the slope of draw duration over the practice session and measured visual recognition learning for the symbols using an old/new 2-AFC task. Results demonstrated that tract microstructure selectively predicted learning outcomes, with left hemisphere pArc and SLF3 tracts predicting drawing learning and the left hemisphere MDLFspl predicting visual recognition learning. These results were replicated using repeat, held-out data and supported with complementary analyses. Results suggest that individual differences in the microstructure of human white matter tracts may be selectively related to future learning outcomes.
Keyphrases
  • white matter
  • multiple sclerosis
  • machine learning
  • healthcare
  • primary care
  • metabolic syndrome
  • young adults
  • mass spectrometry
  • high intensity
  • functional connectivity
  • working memory
  • induced pluripotent stem cells