TAYLOR-IIWO: TAYLOR IMPROVED INVASIVE WEED OPTIMIZATION-ENABLED DEEP QUANTUM NEURAL NETWORK FOR BRAIN TUMOR SEGMENTATION AND CLASSIFICATION

Abstract

The brain tumor classification acts as a fundamental part in the medical areas for diagnosing the brain tumors accurately. The early identification of the brain tumor helps in saving the life of the patients by offering appropriate treatment. This paper devises an approach for the classification of brain tumor using a novel Taylor Improved Invasive Weed Optimization-enabled Deep Quantum Neural Network (Taylor-IIWO-enabled Deep QNN). The pre-processing is the first phase in which the input images are pre-processed, and then the U-Net model is utilized for tumor segmentation. In addition, the extraction of features, such as statistical, Discrete Wavelet Transform (DWT), and shape features, is done in the feature extraction phase. Then, the brain tumor is classified using the Deep QNN classifier, in which the training process is achieved using the devised Taylor-IIWO. The Taylor series and the IIWO are integrated to obtain the developed Taylor-IIWO. In addition, the developed Taylor-IIWO-based Deep QNN has higher accuracy of 0.964, higher sensitivity of 0.967, and higher specificity of 0.983.