Advances in Public Healthcare: Computational Model for Brain Tumour

Abstract

Accurately assessing the size of brain tumours and scheduling for their treatment both heavily depend on determining the overall size of the tumour. For brain tumour, MRI (magnetic resonance imaging) has become the gold standard for diagnosis. Classifying a brain tumour physically takes a lot of time and is mostly dependent on the operator's experience. This work advances public health by presenting a computer approach for brain tumour detection. For the efficient diagnosis of brain tumours, a novel puzzle optimization-driven kernel optimised resnet (PO-KOResnet) technique is put forth. The PO approach is used to maximise the performance of the KOResnet technique. This study used a publicly available data set that was taken from the Kaggle website. It contained 1,500 [256 x 256] MRI pictures, both with and without brain tumours, to evaluate the detection performance of the suggested method. Median filter (MF) is used to enhance the image processing by eliminating noises from the collected raw images. Also, the purpose of this study is to evaluate the suggested method's performance using a variety of measures using the Python platform. The PO-KOResnet technique outperformed other approaches in terms of accuracy value when it came to improving public health consequences globally.