IoT Based Non-Invasive Vital Signs Monitoring in Neonatal Intensive Care Unit (NICU)

Preterm birth is the second major cause of the death of children under five years and the single most cause of death of children within the first month. Such infants need special care and medical supervision until they become strong. Several vital monitoring equipment involves adhesive transducers or electrodes to be directly attached to the skin. As the skin of preterm children is very fragile, the use of conventional technologies leads to even more inconvenience. Also for babies with critical conditions, a strict monitoring system with quick response and data storing capability for future reference is required. Therefore, here we propose a whole automated system in which vital signs of the child will be found without any physical contact using Photo-plethysmogram signal captured by a camera connected to Neonatal Intensive Care Unit. Photo-plethysmogram signals contain information about oxygen and hemoglobin estimated using proposed novel algorithms with the highest possible accuracy. The model is deployed in a Raspberry pi3, which determines the vital signs like heart rate, respiration rate, and spO2 level by extracting iPPG signals from the forehead of the baby. The readings of these vital signs are then transferred to Thingspeak (cloud) and then to user dashboard. The proposed method is non-invasive and portable and is therefore easily accessible across resource-constrained environments.