Effect of Musical Training on the Neural Correlates of Math Processing

Manuscript Published in Neuroscience Letters

Children with severe sensorineural hearing loss now have new possibilities for improvement in their ability to hear and communicate with new technology called Cochlear Implantation. However, not all children experience the same benefit from Cochlear Implantation, with the outcome often being unpredictable, especially in children born deaf, or in whom hearing loss occurred at a very young age. We are investigating the use of functional MRI in children with profound sensorineural hearing loss to determine if residual auditory cortical activity may be observable. Our hope is to use fMRI to guide decision making about cochlear implantation in children.


The effect of formal musical training on separate cognitive domains such as language, mathematical, and spatial continues to be a subject of debate, with some behavioral studies reporting positive associations. Some neuroimaging studies have also demonstrated differences between musicians and non-musicians. Increased corpus callosum size has been found in musicians, as well as increased white matter organization in the genu of the corpus callosum and decreased white matter organization in the corticospinal tract.

Based on evidence from previous behavioral and neuroimaging studies, we therefore hypothesized that formal musical training through the developmental period would affect the neural correlates of math processing. We performed a preliminary investigation of this hypothesis using fMRI. Fifteen normal adults were recruited to participate in the study. Seven of the subjects had studied either a musical instrument or voice since early childhood, while eight had not. The subjects were presented with three problems (either addition or subtraction) to perform mentally. The fractions all contained single digit numbers in both numerator and denominator. No specific instructions were given to the subjects on how to perform the task except the visual reminder on a board prior to the scanning session that a/b + c/d = (ad + bc)/bd.


Significantly greater activation was found (Table 1) in regions including the left fusiform gyrus and left prefrontal cortex for musicians; and the right inferior occipital gyrus, left medial occipital gyrus, right orbital gyrus, and left inferior parietal lobule for non-musicians. The increased activation seen in the left fusiform gyrus (Figure 1, top) is likely associated with increased proficiency in the processing of shape information and visual perceptual semantic processing, generated by years of experience reading and interpreting musical notation. This would also account for the decreased activation seen in visual association areas (Figure 1, bottom), which has been related to practice or habituation effects in studies investigating visuospatial working memory, and the presentation of repeated visual stimuli relative to novel objects. An intriguing alternative hypothesis, however, is that the left fusiform gyrus may be involved in processes involving a more "abstract" level of visual form. Thus, musicians would be employing a more "abstract"" representation of numbers and especially fractions. This hypothesis is corroborated by the decreased activation seen in the left inferior parietal lobule, previously implicated in differences in processing strategy during a serial subtraction task. If musicians were using a more 'abstract' representation of fractions, they would be expected to employ rote calculation procedures to a lesser extent. The increased activation in the left prefrontal cortex (BA 46) in musicians also suggests that the hypothesized link between musical training and improved math performance may also possibly be associated with improved performance of semantic working memory, related to improved conflict resolution.

The statistically significant differences found in the neural correlates of math processing between musicians and non-musicians, even with the small sample size used, provide a strong motivation for a future longitudinal study performed on children in order to probe for a possible causal relationship between musical training and any differences in the neural architecture used for math processing, and to associate those differences with any differences seen in math ability.