The role of an ultrasound-responsive injectable piezoelectric hydrogel in promoting nerve regeneration and alleviating neuropathic pain.

Individuals who invest comparable time and effort in music often attain markedly different skill levels (; ; ; ). Understanding the neural basis of this variability is central to identifying neural markers of learning potential. While machine learning (ML) is well suited to uncover such markers (), its application to neuroimaging data has traditionally relied on expert–novice binary classification (; ; ; ), which does not capture individual variability and the continuous spectrum of training experience typical of the general population (; ). This binary approach also introduces a key confound: experts and novices often differ in both accumulated training and aptitude, making it difficult to separate these factors statistically in neural signatures of expertise (). To address these limitations, we applied prediction-based ML to musical training—a well-established model system of neuroplasticity (; ; ) that exhibits continuous variation across non-specialists (; ) and can be measured with validated psychometric instruments (). Neuroimaging studies of musical expertise have largely focused on structural brain changes associated with intensive, prolonged musical training in adults (; ). Typically, these studies compare highly skilled musicians with untrained listeners (so-called non-musicians), with expertise operationalized categorically through years of formal instruction above an arbitrary threshold—for example, 5, 6, or 10 years (; ; ). This work has consistently revealed large-scale structural differences between groups, primarily within sensorimotor and executive-control networks (; ). These adaptations include changes in gray-matter morphology (; ; ; ; ; ; ; ; ) and white-matter architecture (; ; ; ; ; ; ). This literature showed a predominant “cortico-centric” focus. As detailed in , the majority of diffusion MRI studies () looking at expertise-related white-matter adaptations have deliberately targeted, through a priori region-of-interest analyses, tracts connecting cortical regions: the corpus callosum (; ; ; ; ; ), the arcuate fasciculus (; ; , ), and the superior longitudinal fasciculus (; ). Cortico-subcortical tracts have received far less targeted attention: descending motor pathways such as the corticospinal tract have been examined as primary ROIs in only a handful of studies (; ; ; ), with inconsistencies reported across studies in FA directions (; ): lower for musicians in some (; ) and higher in others (; ; ). Subcortical structures (thalamus, putamen, cerebellum) also show morphological differences (; ; ), but their long-range connectivity patterns and relationship to individual variability in skill acquisition remain less explored. Nevertheless, these studies provide robust evidence that extensive practice is associated with large-scale changes in brain structure and that magnetic resonance imaging (MRI) can detect neural signatures of expertise. However, it remains unresolved whether these signatures extend to the graded variation in training observed in adults with little or no musical training () and whether observed brain–behavior associations primarily reflect accumulated experience, predispositions, third variables, or their combination. Historically, studies have treated non-musicians as a uniform comparison group, conflating the absence of formal instruction with low musical ability (; ; ). In reality, untrained listeners have vast differences in musical experience and how deeply they engage with music (). Also, they span a wide range of musical ability shaped by genetic predispositions and informal musical experience (; ; ; ; ), with some individuals possessing perceptual expertise (‘listening expertise’) comparable with that of trained musicians (). Informal engagement can refine perceptual skills through everyday musical activities (e.g., listening, singing, and dancing), even without formal instruction (). Rather than being a nuisance, this heterogeneity offers a methodological opportunity to disentangle the effects of training and aptitude. In expert–novice comparisons, training is tightly coupled to pre-existing ability, making it difficult to separate experience-dependent plasticity from predisposition (). Because professional musicians typically possess both high perceptual abilities and extensive training, observed neural differences may reflect both pre-existing biological markers of musicality and plasticity (). In contrast, in non-musicians, these factors can be partially decoupled: high aptitude can occur with low or modest training, and training can accumulate with low or modest aptitude (; ). This partial decoupling allows training and aptitude to be examined as partially independent sources of variation in brain structure.