Hybrid Edge-Artificial Intelligence Model for Identification and Classification of Brain Tumours from Computed Tomography Scans

Profound diagnosis of Computed Tomography (CT) scans play a vital role in medical diagnosis among which brain scans help in identifying several neurological problems. We propose a solution by implementing a hybrid edge AI model to identify and classify brain tumours from CT scans. The deep learning architecture developed in the current research is a combination of Convolutional Neural Network(CNN)-based "Mobile Net" architecture integrated with eXtreme Gradient Boost (XGboost) architecture. While traditional computation standards rely on mobile sensing and cloud computing, edge AI implementation on portable devices provides numerous advantages such as quick response time and enhanced data privacy.