Deep Learning-Based Intelligent Model for Neonatal Cyanosis Identification Using SSGNN

Neonatal Cyanosis is a term for the skin's bluish-purple color. Anywhere the skin is tinny, such as the lips, mouth, earlobes, and nails, it is most visible. Cyanosis proposes that the oxygen linked to red blood cells in the bloodstream may have diminished. It can point to a heart or lung issue. The suggested computer vision and deep learning-based intelligent model uses spectral and spatial graph neural networks (SSGNN) applied to facial and body images from NICU. To identify spectral and spatial correlations in the images, the model makes use of deep learning and graph neural networks. The model seeks to learn representations that emphasize traits suggestive of cyanosis by examining the spatial arrangement and spectral properties of pixels in the images. The model is trained and tested using a sizable and varied dataset of neonatal pictures, comprising both cyanotic and non-cyanotic instances. This model is successful, it will be helpful for healthcare workers to identify and diagnose neonatal cyanosis earlier, leading to better healthcare outcomes for newborn infants.