Cloud-Based Framework for Pain Scale Assessment in NICU- A Primitive Study with Infant Cries

Pain scale assessment in a typical neonatal intensive care unit (NICU) is quite challenging for the Neonatology Clinical community. Infant cries, a mere indicator of pain has been considered an effective pathological tool to assess the status of the newborns. Especially in a resource-constrained setting, newborn mortality rate is growing exponentially due to lack of facilities and assessment by the specialist. This specific study attempts to propose a cloud-based framework with the quantitative annotation that will connect the clinician, caretaker and the staff nurse to view the current condition of the newborn. Three variants of cry recording hunger, pain and discomfort were used for this primitive study. Initially, a graphical user interface (GUI) was developed to assess and classify the crying pattern. Power spectral feature and Support Vector Machine (SVM) based neural network were considered for pattern classification. Later using the thingspeak IOT cloud platform, communication was established to clinician/caretaker through the mobile device. A centralized station with the cloud framework facility is currently under development. This primitive study thus enhances the cyber-physical structure to assess the pain scale of newborns and will help in reducing the mortality rate.