Continuous monitoring of temporal skills during long-term in-home training by cochlear implant users

Cochlear implants improve auditory function in individuals with severe hearing loss, yet cochlear implant users often struggle with tasks such as identifying speaker characteristics and musical elements. While music therapy shows promise in addressing these deficits, standardized rehabilitation protocols, especially those focusing on music-based sound recognition, remain underdeveloped. This study evaluates the efficacy of EARPLANTED, a free platform developed at the Faculty of Physics, University of Bialystok, which includes a melodic contour identification test accessible on personal computers and mobile devices. Fifty cochlear implant users and 45 normally hearing volunteers participated in repeated testing sessions. Logistic regression with restricted cubic splines was used to analyse temporal changes in musical perception, offering insights into score distributions and progression over time. Results revealed significant differences in scores, with volunteers outperforming cochlear implant users. However, many volunteers also found the test challenging, highlighting its complexity. Temporal analysis showed that extended use of the platform generally led to improved scores among cochlear implant users. The findings underscore the need for cautious interpretation of melodic contour identification test results, given the difficulty experienced by both cochlear implant users and normally hearing individuals. The platform's utility, particularly for unilateral users, suggests its potential role in auditory rehabilitation. This study highlights the benefits of continuous monitoring and remote data collection via smartphones, providing clinicians and researchers with a practical tool for long-term rehabilitation strategies.