Integrating Computational Modelling into the Ecosystem of Cochlear Implantation: Advancing Access to Diagnostics, Decision-Making, and Post-Implantation Outcomes on a Global Scale.

Disabling hearing loss affects more than 5% of the global population, with numbers expected to double by 2050. The burden is especially high in low- and middle-income countries, where access to cochlear implant (CI) technology and the required follow-up care is limited. While CIs are a proven treatment for certain types of hearing loss, their adoption in these countries is hindered by high costs, the need for specialised rehabilitation, and the financial and time commitment required for long-term device maintenance. Although remote programming has improved accessibility to standard care, specialised interventions for complications remain restricted mainly to areas with clinical centres. Computational modelling offers a promising solution to this access-to-care dilemma. The models may be used to simulate complications, such as non-auditory stimulation (NAS), to investigate and plan personalised interventions, and ultimately predict device parameters, without requiring the recipient's physical presence. Both phenomenological and biophysical models have already demonstrated useful application in CIs: the former streamlines clinical workflows and aims to establish consistency in device fitting, and the latter provides insights into patient-specific auditory biophysiology. Despite decades of research, clinical translation of biophysical models has been limited by data constraints, parameter uncertainty, and validation challenges. In this perspective piece, it is argued that biophysical models have now reached sufficient maturity to be integrated into routine CI care. Apart from the advantages that this approach will bring to the overall advancement of person-centred CI care, it is envisioned to improve accessibility, personalisation, and long-term outcomes for CI recipients in low- and middle-income countries.