A Novel Quantum States Based AI Learning on MRI Brain Slices for the Intracranial Tumor Detection

A tumor in the intra-cranial structure can initiate from various parts of the body thereby proliferating into the cranium. Otherwise, it can originate from the brain cells causing several neurological disorders. Early detection of the tumor promotes effective healing on treatment towards the cyst. The radiotherapist may misinterpret recordings due to various reasons including incorrect segmentation of the MRI image. This proposal works uniquely in brain tumor detection more precisely by the quantum principle implementation on to the pre-processed anisotropically diffused slice. The qubits of the resultant image primarily undergoes image segmentation by an unsupervised learning through Artificial Intelligence (AI) by the implementation of K - Means Clustering (KMC) algorithm. This scales the image qubits and segregates accurate qubits in differentiated quantum states or clusters. One of the state guarantees the extraction of qubits in the Region Of Interest (ROI) accurately by the secondary segmentation through Global Thresholding Methodology (GTM). Thus, by pairing the primary and secondary segmentation (KMC-GTM) promotes AI learning though automated quantitative qubit feature analysis with respect to manually delineated ground truth image. The statistical determination validates the efficacy of the proposal with regard to several segmentation scores along with accuracy of the ROI extraction being 98.93%.