Our Approach
Music Therapy Research
Chronic hemiplegia following stroke is traditionally conceptualized as a deficit of strength, motor control, or cortical representation. While these dimensions are essential, they incompletely explain hallmark features of impairment such as spasticity, impaired initiation, and reduced movement fluidity. At the neurophysiological level, hemiplegia is associated with disrupted inhibitory control, altered corticospinal excitability, and atypical oscillatory activity within motor networks.
We utilize rhythmic stimulation and melodic interventions to enhance motor and cognitive function and emotional well-being in our research participants.
We propose that highly structured musical polyphony, exemplified by the works of J.S. Bach, can serve as an external temporal scaffold, facilitating multi-scale synchronization between neural oscillations, biomechanical pendular dynamics, and group motor behaviour
Balance Training Research
Objective: To propose and evaluate the feasibility of a Distributed Stability Engineering (DSE) framework: a six-node, networked balance rehabilitation system in which patients interact through a shared adaptive state variable governed by distributed control principles.
Methods: We describe the theoretical basis of DSE grounded in distributed dynamics, multi agent reinforcement learning, and ecological coordination theory. A full hardware and software system architecture is presented, including real-time center-of-pressure (COP) acquisition, adaptive coupling algorithms, shared physics modeling, and therapist mediated perturbation control. Novel collective outcome metrics are defined.
Results (Conceptual Feasibility): Existing force plate technology, low-latency networking, and multiplayer virtual environments are sufficient to implement the proposed system. Safety architecture including fall-arrest harnessing and automatic stabilization thresholds ensures clinical viability. The distributed coupling model permits heterogeneous patient inclusion within a shared adaptive framework.
Conclusion: DSE reframes balance rehabilitation from individual stabilization toward adaptive system design. The approach is technically feasible and introduces quantifiable collective metrics that may better reflect real-world stability demands. Empirical validation through controlled trials is warranted.
Background: Conventional balance rehabilitation treats postural instability as an individual deficit and evaluates progress primarily through sway reduction metrics. However, real world balance occurs within dynamic, multi-agent environments in which stability emerges relationally rather than in isolation.
Preclinical signature of Ageing and Geriatric Decline through Virtual Humans Twins (PRESAGE-VHT)
Europe needs tools that detect ageing-related decline early enough to act. On 1 January 2025, the EU population was estimated at 450.6 million and 22.0% was aged 65 years or over, approximately 99 million citizens. In the WHO European Region, the population aged 60+ is projected to increase from 215 million in 2021 to 247 million by 2030 and to over 300 million by 2050. In SHARE-based European data, pre frailty affected 42.9% of older adults and frailty 7.7%. These figures imply a large, clinically important prevention window before overt frailty, dependency and acute service use emerge. The current care system detects much decline too late. Conventional geriatric assessment, frailty scores and electronic frailty indices are clinically valuable, but they are episodic, often descriptive and insufficiently sensitive to early loss of adaptive capacity. Patients may cross a physiological tipping point before their function, cognition, gait or service use visibly changes. PRESAGE-VHT reframes frailty as a late manifestation of earlier multi-system resilience loss. It focuses on the hidden transition in which mental, neuromotor, autonomic, inflammatory/metabolic, sleep and recovery dynamics begin to destabilise. The ambition is not to create another frailty score. It is to build a VHT that models the trajectory towards decline and identifies actionable prevention windows.
Advanced Balance Training Techniques
Rehabilitation programs typically follow a structured and linear progression: patients transition from having their eyes open to closing them, then move onto unstable surfaces, and finally engage in dual-task scenarios. This stepwise increase in complexity is designed to foster gradual improvement. However, the reality of biological learning is rarely so straightforward. In fact, multi-agent learning systems often necessitate recurrent testing and iterative optimization to adapt effectively to the nuances of individual circumstances. When environments shift—whether due to a change in surroundings or a patient's condition— existing policies can degrade, revealing that adaptation is a cyclical process rather than a sequential one
What if balance is not something a person possesses? What if it is something that emerges between people?
For decades, balance rehabilitation has treated instability as an individual deficit. Weak ankles, delayed reflexes, impaired vestibular input—these have been framed as localized failures within a single body. The patient stands alone on a platform, attempting to reduce sway in isolation, and progress is measured by stillness. Less oscillation. Fewer corrections. More control.
But real-world balance is never isolated. The bus accelerates. The crowd shifts. The floor vibrates. Other bodies move unpredictably. Stability in daily life is not achieved in controlled silence but negotiated within dynamic environments populated by other agents. Balance is not an internal possession; it is participation in a living stability ecosystem. Yet our rehabilitation
models remain designed for individuals standing alone.
Traditional therapy isolates patients to minimize variability and unpredictability. It reduces environmental complexity to stabilize performance. It evaluates outcomes through individual sway metrics. But the collective model amplifies structured unpredictability. It introduces relational perturbation. It demands implicit modelling of others’ behaviour. It requires rapid recalibration when another agent shifts strategy. In such a system, the meaningful variables are no longer sway amplitude alone. What matters is convergence speed, robustness to perturbation, recovery after destabilization, synchronization dynamics, and the resilience of the entire network under fluctuating conditions.
The question is not “How still can this person stand?” but “How effectively can this system reorganize under shared instability?” Instability itself changes meaning within this framework. In isolated training, instability is treated as failure. In distributed coordination training, instability becomes curriculum. Oscillation forces exploration. Mismatched responses reveal coupling inefficiencies. Shared error becomes shared information. Learning is driven not by suppressing variability but by intelligently navigating it. This aligns closely with principles from distributed control theory and multi-agent reinforcement learning, where convergence arises from decentralized policy updates within shared environments. But here, the learning is embodied. It is sensorimotor, relational, and physical. The system adapts not through abstract code but through muscle tone, predictive timing, and implicit negotiation.
The implications are radical. Balance rehabilitation should not prepare patients merely to stand quietly in controlled clinical spaces. It should prepare them to function within dynamic, interactive, unpredictable environments. Because outside the clinic, stability is never solitary. We stabilize ourselves in moving crowds, on vibrating transport systems, on uneven terrain, under cognitive load, within social interaction. Real-world equilibrium is participatory.
In this expanded perspective, the therapist’s role transforms. The therapist is no longer primarily correcting ankle strategies or cueing posture. Instead, the therapist becomes an architect of coupling conditions, a designer of perturbation landscapes, a regulator of task constraints, and a shaper of reward structures. Rehabilitation becomes system design rather than symptom correction.
The Art of Sound:
POLYPHONY AS A TEMPLATE FOR NEURO RHYTHMIC RESONANCE IN HEMIPLEGIA
Contemporary neurorehabilitation for hemiplegia has achieved meaningful progress through task-oriented practice, constraint-induced movement therapy, and biomechanical repetition (Winstein et al., 2016). Nevertheless, persistent challenges remain in addressing spasticity, impaired motor timing, and the cognitive and emotional fatigue associated with prolonged individual rehabilitation (Ward, 2017). Converging evidence from systems neuroscience, music cognition, motor control, and social neuroscience suggests that motor recovery is not solely a matter of muscular strength or cortical reactivation, but also of temporal coordination across neural, mechanical, and interpersonal systems (Buzsáki, 2006; Thaut & Hoemberg, 2014).
We introduce Neuro-Rhythmic Resonance (NRR) as a theoretical framework that integrates established findings in rhythmic auditory stimulation, neural entrainment, mirror-neuron–mediated social coupling, and limb biomechanics within a unified, resonance-based model of motor recovery (Rizzolatti & Craighero, 2004; Thaut et al., 1997, Garrison, K.A et al, 2010).
We propose that highly structured musical polyphony, exemplified by the works of J.S. Bach, can serve as an external temporal scaffold, facilitating multi-scale synchronization between neural oscillations, biomechanical pendular dynamics, and group motor behavior (Large & Snyder, 2009).
Rather than presenting NRR as a validated clinical intervention, our approach articulates its conceptual architecture, clarifies its heuristic mechanisms, and outlines testable predictions intended to guide empirical research. By reframing hemiplegic rehabilitation as a problem of biological timing and coordination, this framework aims to catalyze interdisciplinary collaboration across neurorehabilitation, music therapy, systems neuroscience, and physics-informed modeling.
PRESAGE-VHT
PRESAGE-VHT develops a multi-organ, neuro-physiological Virtual Human Twin to detect silent mental, neuromotor and physiological decline in older adults before overt frailty, falls, acute deterioration or loss of independence. Its distinctive innovation is the Resilience Equilibrium Score (RES): a formally defined, uncertainty-aware latent resilience state derived from fractal/multifractal physiology, recovery dynamics and multi-system geriatric data, embedded in a clinically validated VHT.
PRESAGE-VHT will use a hybrid architecture. It is not a black-box risk score. The model couples organ/system layers through a network physiology graph, updates patient-specific latent states over time, and uses interpretable AI to learn non-linear interactions where mechanistic equations are incomplete.
-
Bayesian state-space layer: estimates the time-evolving RES and its credible interval from noisy longitudinal measurements.
-
Network physiology graph: nodes represent neurocognitive/mood, neuromotor, autonomic/cardiovascular, metabolic/renal/inflammatory, medication/nutrition and social-context domains; edges represent coupling strength and cross-system synchrony.
-
Mechanistic submodels/priors: encode known physiological relationships, for example orthostatic recovery, HRV/autonomic regulation, gait-energy demand, inflammatory/metabolic load and medication burden.
-
Interpretable AI layer: learns high-dimensional interactions and provides calibrated risk, uncertainty, driver decomposition and counterfactual “what may change if action is taken” scenarios.
-
Clinical interface layer: converts model output into understandable actions and safety checks rather than unexplained alerts.
Reimagining the Future of Aging: The Hybrid Metaverse Frailty Clinic (HMFC)
The Hybrid Metaverse Frailty Clinic (HMFC) is an innovative healthcare concept designed to transform the prevention, assessment, and management of frailty in older adults. By combining a physical frailty clinic with virtual care environments, telehealth, AI-supported diagnostics, wearable monitoring, digital twins, and community-based support, HMFC offers a more accessible, personalized, and proactive model of care. The platform aims to improve functional independence, reduce avoidable hospitalizations, strengthen patient and caregiver engagement, and support aging well at home and in the community.
Built through collaboration between academia, industry, healthcare professionals, and user groups, HMFC also serves as a research and innovation hub for frailty management. Its approach integrates comprehensive geriatric assessment, preventive care, rehabilitation, remote monitoring, caregiver education, and metaverse-based engagement tools into one connected ecosystem. The goal is to create a sustainable, patient-centered healthcare model that improves quality of life for frail older adults while advancing the future of integrated elderly care.
JOIN THE WALLIS Frailty Commons Think Tank
Wallis Frailty Commons is a proposed public-interest and venture-enabled platform for Crans-Montana and Canton Valais/Wallis. Its purpose is to help older citizens remain independent for longer, to support families and informal caregivers, and to create a measurable alternative to the expensive default path of crisis care, hospital dependency and premature institutionalization.
The concept is derived from the Hybrid Metaverse Frailty Clinic and HMF Care Hub Wallis initiative, which already sets out a powerful human vision: integrative frailty care, comprehensive geriatric assessment, functional independence, prevention, hybrid physical/digital services, research, and community participation. This prospectus reframes that vision into a more realistic and investable business model: asset-light, scalable, strongly governed, and aligned with the political priorities of Wallis.
The decisive shift is to separate clinical care from the broader social, preventive and community support infrastructure. Medical acts remain within existing regulated care pathways when eligible. The new value is created around early detection, care navigation, family support, informal assistance, local service coordination, and data driven prevention - areas that are essential for frailty management but are poorly financed when every need is forced into the traditional medical insurance logic.