Abstract:
Musicians represent an interesting study population for the neurosciences, since they begin their musical training at an early age and dedicate a life-time to the deliberate practice of highly specific skills. The resulting level of excellence makes them an ideal group for investigating a) experience dependent neural plasticity and b) the neural networks underlying those skills. In instrumentalists, several studies have demonstrated plausible experience-dependent functional and structural adaptations of the brain (e.g. cortical expansion representing hands and fingers). However, it remains a matter of debate if similar training related changes might also exist in professionally trained classical singers. Knowledge about the neural networks governing vocal control in experience singers is absent. The present dissertation describes the results from two neuroimaging studies with professionally trained classical singers. The first study investigated the neural networks underlying overt and imagined singing. The second study focused on experience dependent neural plasticity in singers with varying degrees of singing expertise.
Overt singing was found to involve a large bilateral network, focused on areas related to sensorimotor control. Typical language related areas (e.g. Broca and Wernicke) were active during singing, along with their right hemispheric homologue areas. In addition, increased activation comprised the basal ganglia, the thalamus, the cerebellum, the anterior cingulate cortex, and the bilateral insula. In contrast to previous studies with instrumentalists, mental singing also activated the primary and secondary sensorimotor areas. In agreement with previous studies, imagery did not activate the primary auditory cortex. Mental singing differed from overt singing in activation of a large fronto-parietal network, and increased involvement of areas associated with emotion processing.
In the second, cross-sectional study, experience-dependent changes in neural networks were investigated in subjects with varying singing proficiency: students with no professional singing experience, professionally trained vocal students, and professional opera singers. These groups differed with respect to their level of singing skills but also in their amount of accumulated singing practice. Comparisons between vocal students and laymen revealed increased activation of bilateral somatosensory cortex (S1) in the representation of larynx and articulators. No difference was observed in primary motor cortex (M1), which stands in contrast to previous results in instrumentalists. Further experience-dependent activation was found within inferior parietal cortex (IPC), dorsolateral prefrontal cortex (DLPFC), the pallidum, and the cerebellum. The comparison opera singers versus laymen showed a similar pattern, comprising S1, the basal ganglia, IPC, cerebellum, and the thalamus. In addition, increased activity involved also right M1 in a circumscribed area. Increased right S1/M1 was also found when opera singers were compared to vocal students, together with the putamen, IPC, and the cerebellum.
A regression analysis based on the accumulated singing practice confirmed that the amount of dedicated singing practice correlates positively with activation in bilateral S1, DLPFC, IPC, and basal ganglia.