An Efficient Hybrid Classifier for MRI Brain Images Classification Using Machine Learning Based Naive Bayes Algorithm

• Published in: SN computer science Vol. 4; no. 3; p. 223
• Main Authors: Nayak, Madhu M., Kengeri Anjanappa, Sumithra Devi
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.05.2023; Springer Nature B.V
• Subjects: Accuracy; Advances in Computational Intelligence for Artificial Intelligence; Algorithms; Bayesian analysis; Body parts; Brain; Brain cancer; Classification; Classifiers; Computer Imaging; Computer Science; Computer Systems Organization and Communication Networks; Data Structures and Information Theory; Datasets; Discriminant analysis; Feature extraction; Image classification; Image enhancement; Image quality; Information Systems and Communication Service; Internet of Things and Data Analytics; Machine Learning; Magnetic resonance imaging; Mathematical models; Medical imaging; Medical science; Neural networks; Neuroimaging; Noise reduction; Original Research; Parameter identification; Pattern Recognition and Graphics; Software Engineering/Programming and Operating Systems; Splitting; Support vector machines; Tomography; Tumors; Vision; Wavelet transforms; Magnetic resonance imaging (MRI); Positron emanation tomography (PET); Random subspace (RS); Bayesian Network (BN); Random forest (RF); Feed forward-artificial neural network (FF-ANN)
• ISSN: 2661-8907; 2662-995X; 2661-8907
• DOI: 10.1007/s42979-022-01614-y

In recent days, advanced techniques are used to compare the analysis of medical images, identifying, pre-processing and interpreting the images. As a result, visualizing of images have greatly diversified in medical sciences domain. The magnetic resonance imaging (MRI) scanner is commonly used to identify and differentiate between normal and abnormal medical images. But in recent days, analysis of new variant of diseases is very difficult, hence adopting new advanced techniques in data pre-processing and analysis of medical images is very essential. This paper proposes a hybrid naive-Bayes classifier for MRI brain image differentiation process. The image quality of human body parts is enhanced including the brain images to improve classification accuracy rate by efficiently differentiating normal and abnormal images containing the disorders and injuries using a hybrid naive-based classifier in MRI brain images. The image pre-processing, feature extraction and noise reduction is achieved using proposed model. The proposed model is processed in four different steps such as image pre-processing, feature extraction and feature reduction using a naive Bayes classifier. The median filter is effectively utilized by a hybrid algorithm to remove noise such as scalp and skull. The performance analysis has been conducted by collecting a huge number of sample images and efficiently differentiating normal and abnormal images using the proposed algorithm. The comparative analysis has been conducted between proposed algorithm with existing methods like Random subspace with random forest (RS with RF), random subspace with Bayesian Network (RS with BN) and Feed Forward-ANN (FF-ANN). The aim of this work is to improve the classification accuracy with efficient and fast method which identifies the small number set of optimal parameters. The main purpose of the proposed mathematical model is to increase the accuracy rate of normal image classification and abnormal image classification with respect to classification methods like RS with RF, RS with BN and FF-ANN. The proposed hybrid naive Bayes classifier gives a 35–65% splitting ratio for training and splitting ratio. With respect to improvements in normal and abnormal classification of an image, samples are 2%, 3%, and 2.5% using methods (RS with RF), (RS with BN) and feed forward-ANN (FF-ANN), respectively.