Ensemble music-making constitutes a temporally structured, multimodal environment with differentiated roles and hierarchical coordination (9, 10). Computational and theoretical models suggest engagement of predictive coding for social synchronization and sensorimotor coupling (8, 21). In typically developing populations, activity spans fronto-striatal, temporo-parietal, limbic, and brainstem circuits contributing to entrainment and alignment (5–7, 22).

Structured music training has been associated with experience-dependent plasticity and enriched-environment effects across motor, cognitive, and socio-affective systems (4, 10, 12, 23, 24). Reported neurostructural adaptations include enhanced interhemispheric connectivity, reorganization of motor regions, and strengthened audio–motor coupling in pediatric and adult cohorts (7, 25, 26). These processes are salient during the 9–25-year developmental window, when sensorimotor, executive, and socio-affective systems show heightened susceptibility to experience-driven modulation (27–29).

Overall, these findings support the hypothesis that clinically adapted orchestral practice could function as a multimodal enriched environment during sensitive developmental windows. Within this perspective, I-SOUND is introduced as an exploratory framework to examine such hypotheses in NDD through sequential individual–ensemble modules and regulatory transfer mechanisms aimed at ecological validity and reproducibility (15, 16).

Evidence on neural substrates of musical interaction in NDD is limited; most data derive from neurotypical or mixed samples. Neuroimaging and electrophysiology suggest that ensemble music can enhance audio–motor coupling, engage mirror neuron systems, and modulate dopaminergic and serotonergic pathways, although current evidence derives predominantly from cross-sectional studies and remains preliminary (4–7). Predictive coding accounts posit minimization of prediction errors in melody, rhythm, and harmony, potentially improving synchrony (8). Groove-rich music and salient visual cues may promote motor engagement and alignment (21), recruiting fronto-striatal circuits, insula, and anterior cingulate cortex—regions implicated in timing, emotion, and social regulation. Group-based musical practices have also been associated with changes in neurotrophic and stress-related indices, but these findings remain exploratory and are not the focus of the present program, which prioritizes clinically validated functional outcomes. These mechanisms remain plausible targets requiring structured testing in NDD, which I-SOUND is designed to explore.

Several ensemble programs were conceived primarily for psychosocial inclusion rather than as trial-ready clinical protocols; their evidence base is largely observational or pilot-level, including El Sistema (22, 30–33), Nordoff-Robbins (34), Strokestra (35, 36), Esagramma (37), AllegroModerato (38), and community-oriented frameworks such as Community Music Therapy (39–45). These initiatives prioritize expressive and relational aims and provide useful observations—for example, groove-rich repertoires and salient visual interaction may support coordination and prosocial behaviors (21). Yet most lack standardized eligibility criteria, fidelity thresholds, prespecified endpoints, or multisite procedures with blinded assessment in pediatric or transitional-age NDD cohorts (46, 47).

In contrast, I-SOUND may be described as a clinical framework that proposes to translate these inclusive premises into a trial-ready structure, organized around the modular sequence IMT-OMT-MIT-P. Its design is intended to align with DSM-5-TR/CDDR diagnostic criteria, to incorporate predefined fidelity metrics with an intraclass correlation coefficient (ICC) ≥ 0.80, and to follow a two-phase methodology (Phase 1 STROBE; Phase 2 CONSORT) aimed at enhancing transparency, reproducibility, and clinical applicability. In this context, I-SOUND is introduced as a tentative clinical framework differing by: (i) a standardized IMT–OMT sequence regulated by MIT-P; (ii) TIDieR-compliant specification with fidelity thresholds (ICC ≥ 0.80) and replication materials; (iii) a two-phase design (Phase 1 STROBE; Phase 2 CONSORT) with stratified randomization and blinded outcomes; and (iv) multicenter feasibility through predefined adaptations and accredited provider training (20, 46, 47).

Music-based interventions (MBI) in NDD have reported effects in motor, language, and socio-emotional domains, particularly in autism spectrum disorder (ASD), CP, and selected genetic syndromes (46, 48–50). Modalities include Rhythmic Auditory Stimulation (RAS), therapeutic singing, and instrumental training, with substantial heterogeneity in dosage, duration, and complexity. Reviews indicate that music may support functional and structural adaptations, and that ensemble formats could contribute to socially mediated plasticity (10, 23). Nonetheless, most investigations are modality-specific and seldom integrate individualized and collective modules within stratified cohorts. No controlled studies have validated an orchestral framework explicitly combining IMT, OMT, and MIT-P in NDD, nor examined multisite feasibility with blinded assessment and fidelity monitoring.

Despite growing interest, the MBI literature in NDD is marked by methodological variability, short follow-up, and limited standardization (46, 51, 52). Incorporating predictive-processing models and groove-mediated engagement within structured designs may clarify ensemble-driven plasticity. Yet interventions rarely include predefined fidelity metrics, stratification strategies, or blinded assessments. To our knowledge, no published protocol integrates IMT, OMT, and MIT-P within a unified, developmentally calibrated orchestral framework with systematic clinical monitoring and multicenter evaluation. I-SOUND was therefore developed as a cautious methodological proposal aligned with DSM-5-TR/CDDR, operationalizing neurofunctional premises through MIT-P cycles and sequencing modules to accommodate developmental transitions. Generalization is conceptualized as near and far transfer, operationalized through time-defined MIT-P cycles to support consolidation and cross-context application.

I-SOUND, a clinically intensive extension of the Euterpe Method, is structured according to TIDieR (20) and integrates principles of experience-dependent plasticity (29), stratified functional targeting, and adaptive musical codification (Figure 1). IMT and OMT are derived from the EM Active algorithm, initially designed for pediatric cohorts and later adapted for adolescents and young adults with NDD. EM Active functions as a clinical device supporting adaptive resilience and longitudinal functional targeting, maintaining methodological continuity across developmental phases (15, 16).

I-SOUND is a modular, cyclic framework addressing three recurrent gaps: limited integration of individual and ensemble modalities, absence of a formalized transfer mechanism, and lack of standardized, reproducible procedures. Its architecture aligns with neurofunctional substrates implicated in adaptive plasticity, predictive timing, and cross-modal synchronization (71). The design seeks to balance ethical inclusion with internal validity through adaptive eligibility and implementation strategies. Within this framework, MIT-P coordinates transfer between IMT and OMT via temporally defined practice sequences, supporting learning, consolidation, and fidelity across heterogeneous profiles. Distinct from community-oriented programs, I-SOUND incorporates prespecified fidelity thresholds (e.g., ICC ≥ 0.80), stratified randomization, blinded assessment, multicenter planning, and systematic video coding by multiple trained raters to mitigate observational bias (72).

IMT targets predictive timing and multimodal sensorimotor integration; OMT extends these capacities in structured ensemble contexts. MIT-P ensures inter-modular coherence through repeated cycles (66), with potential improvements in untrained tasks (67), conceptualized as near and far transfer and integrating automatic and reflective components (68–70), hypothetically engaging cortical plasticity and transient disinhibition (24). Collectively, modules could activate bilateral audio–motor networks, including supplementary motor area, premotor cortex, basal ganglia, and cerebellum, consistent with beat-based mechanisms (73, 74). Hierarchical synchronization and role differentiation, interpreted within predictive-coding models, may reduce error signals and support emotional regulation and interpersonal coordination (8). Observations from collective music-making in neurotypical cohorts point to improvements in motor control, emotion regulation, and plasticity, which require empirical verification in NDD. High-groove music and visual social cues have been associated with increased movement energy and coordination in ensembles (21). Plasticity studies suggest that instrumental practice can modulate fronto-parietal and cerebellar activity, with auditory and striatal measures predicting learning rate; preliminary evidence indicates modulation of neurotrophic and stress-related biomarkers, although findings remain exploratory (10, 23, 75, 76). Caregiver involvement, ecological momentary assessment, and wearable sensors are positioned to increase ecological validity, personalize parameters, and enhance sensitivity to change; wearables remain exploratory and excluded from efficacy analyses.

The biphasic design—observational modeling followed by RCT—aims to balance ecological validity with controlled hypothesis testing (77, 78). Strengths include validated multidomain assessments (79), fidelity controls, and modular implementation adaptable to diverse contexts (80). Preliminary Italy–Lebanon experience may support transcultural replication, pending empirical confirmation. The inclusion of EMA-T/EMA-P (Supplementary materials 1, 2) and TIDieR (Supplementary Table 1) contributes to reproducibility, sensitivity to change, and transparency, aligning with STROBE, CONSORT, and TIDieR.

Limitations include the absence of harmonized international protocols (81), resource demands (specialized personnel, adapted instruments, flexible spaces) (82), and the methodological challenge of balancing ethical inclusion with internal validity. Another limitation is the lack of accredited training for the conductor–music therapist role, requiring structured pathways. Current mitigation—modular equipment, inter-institutional collaborations, and training initiatives—remains partial.

Safety is addressed through pre-session health checks, continuous therapist monitoring, and post-session debriefings, with adaptive procedures allowing modification or suspension in cases of instability (78, 83). Incidents are centrally logged and reviewed weekly by the clinical team. To limit bias, systematic video recording with multi-rater cross-coding improves reliability, although resource-intensive. Operational constraints persist (specialized personnel, adapted instruments, institutional variability) (82); current mitigation via modular setups and shared facilities provides only partial relief. Accredited training and harmonized protocols remain prerequisites for scaling.

The modular structure may be adapted to specific NDD subgroups (71, 84). Wearable technologies and longitudinal EMA could enhance ecological monitoring. Given that the current program already includes structured follow-up assessments up to 24 months, subsequent research should investigate whether extending monitoring beyond this timeframe is clinically informative (85). In future applications, if the intervention is requested for periods longer than 20 weeks, additional follow-up assessments will be planned according to treatment course and outcomes observed at earlier timepoints, allowing data-driven adaptation of longitudinal monitoring (86). Furthermore, planned developments include multicenter implementation with harmonized standards, accredited training for the conductor–music therapist role, and a standardized starter-kit (instrument set, compositional/therapeutic library, fidelity manual). International collaborations may facilitate scalability while maintaining cultural adaptability.