Intelligent wearable allows out-of-the-lab tracking of developing motor abilities in infants

This research presents a novel method for objectively assessing infants' motor abilities using a multi-sensor wearable, MAIJU. The study involved 59 infants aged 5-19 months, recorded during spontaneous play. A deep learning-based classifier, trained on a novel gross motor description scheme, achieved human-equivalent accuracy in quantifying postures and movements. Aggregated data were used to create the Baba Infant Motor Score (BIMS), strongly correlating with chronological age (r = 0.89, p < 1e-20). This wearable technology enables out-of-hospital assessment, offering a scalable and objective solution for early neurodevelopmental care.

Manu Airaksinen, Anastasia Gallen, Anna Kivi, Pavithra Vijayakrishnan, Taru Häyrinen, Elina Ilén, Okko Räsänen, Leena M. Haataja, Sampsa Vanhatalo