Computer vision based efficient segmentation and classification of multi brain tumor using computed tomography images

Abstract

This study aims to highlight the effectiveness of computer vision (CV) techniques in classifying brain tumors using a comprehensive dataset consisting of computed tomography (CT) scans. The proposed framework comprises six types of brain tumors, including benign tumors (Meningioma, Schwannoma, and Neurofibromatosis) and malignant tumors (Glioma, Chondrosarcoma, and Chordoma). The acquired images underwent pre-processing steps to enhance the dataset's quality, including noise reduction through median and Gaussian filters and region of interest (ROIs) extraction using an automated binary threshold-based fuzzy c-means segmentation (ABTFCS) approach. A total of 900 CT-scan images were utilized, 150 images per tumor class, each with a size of 512 × 512 pixels, and 4 ROIs taken per image, so the total dataset size is 3600 (900 × 4) attributes. After pre-processing, the dataset was further analysed to extract 135 statistical multi-features for each ROI. An optimized set of 12 statistical multi-features was selected to identify the most relevant features using a feature selection technique based on correlation. For the classification stage, the optimized statistical multi-feature dataset was evaluated using five computer vision classifiers: multilayer perceptron (MLP), BayesNet, PART, random tree, and randomizable filtered classifier, employing a 10-fold cross-validation method. Among these classifiers, MLP with fine-tuned hyperparameters achieved a promising accuracy rate of 97.83%.

Keywords: ABTFCS; Brain tumor; Computer vision; Multilayer perceptron; Optimized statistical multi-features.

Fig. 1

Sample of CT scans of six types of brain tumors were used in this study.

Fig. 2

The proposed framework for the classification of six types of brain tumors based on CT-scan.

Algorithm 1

Proposed algorithm for the classification of brain tumors based of CT Scans.

Fig. 3

Outcomes of the proposed segmentation framework for brain tumor localization in CT scans.

Fig. 4

Fine-tuned MLP-based classification framework for six types of brain tumors using CT scan-derived optimized statistical multi-features.

Fig. 5

Overall accuracy for CV classifiers employed on statistical multi-features Dataset.

Fig. 6

The classification accuracy graph depicts the performance of the MLP classifier on a statistical multi-feature dataset for six different types of brain tumor.

Fig. 7

Overall accuracy for CV classifiers employed on optimized statistical multi-features dataset.

Fig. 8

The classification accuracy graph depicts the performance of the MLP classifier on an optimized statistical multi-feature dataset for six different types of brain tumor.

Fig. 9

The comparative analysis of classification accuracy of statistical multi features and optimized statistical multi features datasets.

Fig. 10

Overall accuracy for CV classifiers employed on hybrid multi-features validation dataset.

Fig. 11

The accuracy graph compares the MLP performance of proposed and the validation optimized hybrid multi-feature dataset for brain tumor classification based on CT-Scan.