f MRI brain tumor classification using HOG features selected via impurity-based importances measure
FOREX Press I. J. of Electrical & Electronics Research
Support Open Access
  • Home/
  • IJEER-120416

Research Article |

MRI Brain Tumor Classification Using HOG Features Selected via Impurity Based Importances Measure

Author(s): Yasser Nizamli* and Anton Filatov

Published In : International Journal of Electrical and Electronics Research (IJEER)        Volume 12, Issue 4

Publisher : FOREX Publication

Published : 30 November 2024

e-ISSN : 2347-470X

Page(s) : 1251-1257

DOI: https://doi.org/10.37391/IJEER.120416

Abstract

MRI is considered the primary method for confirming the diagnosis of brain tumors and choosing the appropriate treatment. Automating the process of detecting brain tumors in MRI images using deep models has become a popular trend in the scientific research community. However, deep neural networks require a large volume of data to avoid overfitting, which is not ideally available. This is where handcrafted features come in handy. In this paper, we present an efficient approach for brain tumor classification that can outperform deep CNN models. In the proposed system, the histogram of oriented gradients algorithm is used to extract feature descriptors from brain MRI images. The extracted features are processed using a random forest algorithm, where each decision tree performs the task of evaluating feature importances via the impurity metric, while all estimators collaborate in selecting the effective feature set. Finally, the Fine KNN algorithm is used to classify the types of brain tumors based on the numerical features obtained. The presented model achieved a high-test accuracy of 99.35% with an F1-score of 99.30%, outperforming many deep models.

Keywords: MRI, Brain tumors, HOG, Random forest, Impurity, Feature importances, Fine KNN.




Yasser Nizamli*, Tishreen University, Lattakia, Syria; Email: yasser.nizamli@tishreen.edu.sy

Anton Filatov, Saint Petersburg Electrotechnical University “LETI”, Saint Petersburg, Russia; Email: yanizamli@stud.etu.ru

    [1] Sowrirajan, S. R.; Balasubramanian, S. Brain tumor classification using machine learning and deep learning algorithms. IJEER 2022, Volume 10, pp. 999–1004.
    [2] Alavi, S. E.; Zare, E.; Rashti, M. j. Brain Tumors Detection on MRI Images through Extracting HOG Features. JOAASR 2018, Volume 2, pp. 9–25.
    [3] Nizamli, Y.; Filatov, A. Improving transfer learning performance for abnormality detection in brain MRI images using feature optimization techniques. 2024 XXVII International Conference on Soft Computing and Measurements (SCM), Saint Petersburg, Russia, 2024, pp. 432–435.
    [4] Biswas, A.; Islam, M. S. A hybrid deep CNN-SVM approach for brain tumor classification. JISEBI 2023, Volume 9, pp. 1–15.
    [5] Jaspin, K.; Selvan, S. Multiclass convolutional neural network based classification for the diagnosis of brain MRI images. Biomedical Signal Processing and Control 2023, Volume 83.
    [6] Khan, S. U. R.; Zhao, M.; Asif, S.; Chen, X. Hybrid-NET: A fusion of DenseNet169 and advanced machine learning classifiers for enhanced brain tumor diagnosis. IMA 2024, Volume 34.
    [7] Rahman, T.; Islam, M. S. MRI brain tumor detection and classification using parallel deep convolutional neural networks. Measurement: Sensors 2023, Volume 26.
    [8] Sowrirajan, S. R.; Balasubramanian, S.; Raj, R. S. P. MRI brain tumor classification using a hybrid VGG16-NADE model. Brazilian Archives of Biology and Technology 2023, Volume 66.
    [9] Passa, R. S.; Nurmaini, S.; Rini, D. P. YOLOv8 based on data augmentation for MRI brain tumor detection. Scientific Journal of Informatics 2023. Volume 10.
    [10] Deepak, S.; Ameer, P. M. Brain tumor categorization from imbalanced MRI dataset using weighted loss and deep feature fusion. Neurocomputing 2023, Volume 520, pp. 94–102.
    [11] Kibriya, H.; Masood, M.; Nawaz, M.; Nazir, T. Multiclass classification of brain tumors using a novel CNN architecture. Multimedia Tools and Applications 2022, Volume 81, pp. 29847–29863.
    [12] Nizamli, Y.; Filatov, A.; Fadel, W.; Shichkina, Yu. Accurate Anomaly Detection in Medical Images using Transfer Learning and Data Optimization: MRI and CT as Case Studies. 2024 V International Conference on Neural Networks and Neurotechnologies (NeuroNT), Saint Petersburg, Russia, 2024, pp. 170–173.
    [13] Brain tumor dataset. Available online: https://figshare.com/articles/dataset/brain_tumor_dataset/1512427/5 (accessed 26.03.2024).
    [14] Zhang, B.; Dong, X.; Hu, Y.; Jiang, X.; Li, G. Classification and prediction of spinal disease based on the SMOTE-RFE-XGBoost model. PeerJ Computer Science 2023, Volume 9.
    [15] Ahn H. Performance evaluation of a feature-importance-based feature selection method for time series prediction. JICCE 2023, Volume 21, pp. 82–89.

Yasser Nizamli and Anton Filatov (2024), MRI brain tumor classification using HOG features selected via impurity-based importances measure. IJEER 12(4), 1251-1257. DOI: 10.37391/IJEER.120416.