A number of methods, like magnetic resonance imaging (MRI), biopsies, CT scans, and others, can be used to find brain tumours (BT). One of the most effective tools available today for identifying BT is MRI. The MRI scan is used to identify BT that are well along in their development. The procedure of segmentation is crucial for removing questionable areas from intricate medical imaging. A beneficial perspective and level of accuracy for initial brain cancer diagnosis can be provided by automatic BT recognition utilising MRI. In this work, we evaluate the research of several Machine Learning (ML) algorithms, including those employed for the early diagnosis of BTwhich are Support Vector Machines (SVM), K-Nearest Neighbour (KNN), Generalised Regression Neural Networks (GRNN), fuzzy C Means (FCM), fuzzy C Means (FCM) with Particle Swarm Optimisation (PSO), SVM-KNN Hybrid Classifier, Probabilistic Neural Network (PNN), 2D CNN & auto-encoder networks, K-Nearest Neighbours (KNN), Multilayer Perceptron (MLP) in which SVM-CNN hybrid shows highest accuracy i.e. 98% and MLP showing lowest accuracy i.e. 28% respectively.
Paper
Full text
Brain Tumour Identification and Prediction with Machine Learning
Semantic Scholar · Medicine · 2023
Abstract
A number of methods, like magnetic resonance imaging (MRI), biopsies, CT scans, and others, can be used to find brain tumours (BT). One of the most effective tools available today for identifying BT is MRI. The MRI scan is used to identify BT that are well along in their development. The procedure of segmentation is crucial for removing questionable areas from intricate medical imaging. A beneficial perspective and level of accuracy for initial brain cancer diagnosis can be provided by automatic BT recognition utilising MRI. In this work, we evaluate the research of several Machine Learning (ML) algorithms, including those employed for the early diagnosis of BTwhich are Support Vector Machines (SVM), K-Nearest Neighbour (KNN), Generalised Regression Neural Networks (GRNN), fuzzy C Means (FCM), fuzzy C Means (FCM) with Particle Swarm Optimisation (PSO), SVM-KNN Hybrid Classifier, Probabilistic Neural Network (PNN), 2D CNN & auto-encoder networks, K-Nearest Neighbours (KNN), Multilayer Perceptron (MLP) in which SVM-CNN hybrid shows highest accuracy i.e. 98% and MLP showing lowest accuracy i.e. 28% respectively.