Brain tumor classification with deep learning: A comparative study of AlexNet and EfficientNetB0 using principal component analysis

Kerem Gencer

doi:10.26900/hsq.2804

Authors

Kerem Gencer Afyon Kocatepe University / Türkiye 0000-0002-2914-1056

10.26900/hsq.2804

Abstract

In this study, we investigate the effectiveness of two advanced deep learning models, namely AlexNet and EfficientNetB0, for brain tumor classification using MRI images. Brain tumors pose a significant medical challenge due to their complex structure and critical impact on brain functions. Accurate detection and classification of brain tumors using medical imaging, especially Magnetic Resonance Imaging (MRI), is crucial for effective diagnosis and treatment planning. Recent advances in deep learning have facilitated the development of sophisticated algorithms that significantly improve the accuracy of tumor detection and classification. AlexNet, known for its deep convolutional layers on the MRI dataset, and EfficientNetB0, which uses a compound scaling method, were trained and evaluated on this dataset. Principal Component Analysis (PCA) was used to reduce the dimensionality of the extracted features, thereby optimizing the performance and computational efficiency of the models. Our results show that the EfficientNetB0 model achieves superior performance, with an accuracy of 97.87%, representing a 17.45 percentage point improvement over the AlexNet model, which achieved 80.42% accuracy. EfficientNetB0 demonstrates higher precision, recall, and F1 scores across all tumor classes, demonstrating its robustness and reliability for clinical applications. The findings of this study highlight the potential of combining advanced deep learning models with PCA to improve the accuracy and efficiency of brain tumor classification. This approach not only improves diagnostic accuracy but also reduces computational complexity, making it applicable for real-world clinical application.

Keywords:

AlexNet brain tumor classification deep learning EfficientNetB0 MRI principal component analysis

Downloads

Download data is not yet available.

References

Cheng J. Brain tumor dataset. Figshare. Dataset, 1512427(5). doi: 10.6084/m9.figshare.1512427.v5.

Khagan A. Figshare Brain Tumor Dataset. Kaggle. Available from: https://www.kaggle.com/datasets/ashkhagan/figshare-brain-tumor-dataset/data. (Accessed on Jan 10, 2025).

Siddique MAB, Khatun S, Alimuzzaman M, Hossain MR, Roy SK, Rahman MM, et al. Deep convolutional neural networks model-based brain tumor detection in brain MRI images. Proc IEEE Int Conf I-SMAC. 2020;4:313-8. doi: 10.1109/I-SMAC49090.2020.9243461.

Akkus Z, Galimzianova A, Hoogi A, Rubin DL, Erickson BJ. Deep learning for brain MRI segmentation: state of the art and future directions. J Digit Imaging. 2017;30:449-59. doi: 10.1007/s10278-017-9983-4.

Pereira S, Pinto A, Alves V, Silva CA. Brain tumor segmentation using convolutional neural networks in MRI images. IEEE Trans Med Imaging. 2016;35:1240-51. doi: 10.1109/TMI.2016.2538465.

Goceri E. Medical image data augmentation: techniques, comparisons and interpretations. Artif Intell Rev. 2023;56:12561-605. doi: 10.1007/s10462-023-10453-z.

Goni MR, Ruhaiyem NIR. Salient feature extraction using attention for brain tumor segmentation. Proc IEEE Int Conf Comput. 2022;7:1-6. doi: 10.1109/ICOCO56118.2022.10031677.

Özbay FA, Özbay E. Brain tumor detection with mRMR-based multimodal fusion of deep learning from MR images using Grad-CAM. Iran J Comput Sci. 2023;6:245-59. doi: 10.1007/s42044-023-00137-w.

Yildirim M, Altun Z, Avcı E, Bayrak S, Cınar Yücel S, Güneş S. Detection and classification of glioma, meningioma, pituitary tumor, and normal in brain magnetic resonance imaging using deep learning-based hybrid model. Iran J Comput Sci. 2023;6:455-64. doi: 10.1007/s42044-023-00139-8.

Reza SM, Mays R, Iftekharuddin KM. Multi-fractal detrended texture feature for brain tumor classification. Proc SPIE Int Soc Opt Eng. 2015;4:905-10. doi: 10.1117/12.2083596.

Hammad M, Ismail M, Geman O, Dobre C. Efficient brain tumor detection with lightweight end-to-end deep learning model. Cancers. 2023;15:2837. doi: 10.3390/cancers15102837.

Gumaei A, Hassan MM, Hassan MR, Alelaiwi A, Fortino G. A hybrid feature extraction method with regularized extreme learning machine for brain tumor classification. IEEE Access. 2019;7:36266-73. doi: 10.1109/ACCESS.2019.2904145.

Ahmadi M, Nasrollahi A, Mohammadzade H, Fakhari A. Detection of brain lesion location in MRI images using convolutional neural network and robust PCA. Int J Neurosci. 2023;133:55-66. doi: 10.1080/00207454.2021.1883602.

Islam MK, Ashraf A, Alam H, Islam S, Rahi KS, Billah K, et al. Brain tumor detection in MR image using superpixels, principal component analysis and template-based K-means clustering algorithm. Mach Learn Appl. 2021;5:100044. doi: 10.1016/j.mlwa.2021.100044.

Saraswathi V, Gupta D. Classification of brain tumor using PCA-RF in MR neurological images. Proc IEEE Int Conf Commun Syst Netw. 2019;11:1-6. doi: 10.1109/COMSNETS.2019.8711010.

Gencer K, Gencer G. Hybrid deep learning approach for brain tumor classification using EfficientNetB0 and novel quantum genetic algorithm. PeerJ Comput Sci. 2025;11:e2556. doi: 10.7717/peerj-cs.2556.

Gencer K, Gencer G, Cizmeci İH. Deep learning approaches for retinal image classification: A comparative study of GoogLeNet and ResNet architectures. Int Sci Vocat Stud J. 2024;8(2):123-8. doi: 10.47897/bilmes.1523768.

Raza A, Ayub H, Khan JA, Ahmad I, Salama AS, Daradkeh YI, et al. A hybrid deep learning-based approach for brain tumor classification. Electronics. 2022;11(7):1146. doi: 10.3390/electronics11071146.

Roy AG, Conjeti S, Navab N, Wachinger C. QuickNAT: A fully convolutional network for quick and accurate segmentation of neuroanatomy. NeuroImage. 2019;186:713-27. doi: 10.1016/j.neuroimage.2018.11.042.

Yoon J, Lee J, Kang J. Efficient deep learning-based classification of OCT images for the diagnosis of retinal diseases. J Healthc Eng. 2020;2020. doi: 10.1155/2020/8025393.

Rajagopalan AV, Karnan M, Menaka R. Automated classification of retinal diseases using deep learning algorithm. Int J Biomed Eng Technol. 2021;37(2):185-98. doi: 10.1504/IJBET.2021.115708.