A Computer Vision Approach for Infant Cognitive Monitoring

Emerging technology in telemedicine has seen significant progress realized in the remote monitoring of patients. However, the focus remains on mechanical measurements such as blood pressure, respiration, or heart rate. Further, the attention resides on physical health rather than cognitive monitoring, rendering them inadequate for monitoring developmental milestones among infants or patients receiving psychomotor therapy. Imaging solutions that monitor brain activity such as Electroencephalography and magnetoencephalography address this, however they require skills and present high costs. This research presents an accessible imaging prototype that automates infant cognitive monitoring using object-tracking techniques. Experimental results show that computer vision can automate the process used by paediatricians for manually monitoring a baby's cognitive development. Limited resources and access to health care in marginalized communities reduce the frequency of postnatal care visits, subsequently increasing infant mortality rates. Hence, the proposed solution supports the global sustainable development goals of promoting healthcare for all.