NeuroScan: An Automated Brain Tumor Detection System Using Machine Learning on MRI Images

Abstract

Brain tumors are amongst the neurological disorders that are of high concern. Early and accurate diagnosis is a high priority for improvement in patient survival rates. Manual MRI image analysis is time-consuming and highly dependent on expert radiologists. In the present work, the authors propose Neuro Scan-a fully automated system for brain tumor detection and classification based on a machine-learning framework. MRI image quality is enhanced by resizing, grayscale conversion, noise reduction, and Min-Max normalization. Next, the features were extracted and classification was performed using SVM and Logistic Regression. The dataset consisted of approximately 2500 brain MRI images, which are collected from publicly available repositories. The experimental results indicated the better performance of the SVM classifier with accuracy up to 93-95% outperforming Logistic Regression. Therefore, this system greatly assists clinicians by offering more efficient, consistent, and high-speed tumor detection, aiding in early diagnosis and treatment planning. The results have demonstrated the efficiency and feasibility of such machine learning-based systems for medical image analysis and clinical decision support.

Introduction

The document focuses on the automatic detection of brain tumors using MRI images, highlighting the limitations of manual diagnosis such as time consumption, subjectivity, and human error. It emphasizes the need for an automated system that can accurately identify tumors using machine learning techniques.

Traditional image processing methods and classical machine learning algorithms like SVM and Logistic Regression have been used for tumor detection, but they often struggle with variations in tumor size, intensity, and MRI image quality. While deep learning methods offer higher accuracy, they require large datasets and high computational resources.

To address these challenges, the proposed system called Neuro Scan is introduced. It uses a machine learning-based pipeline involving image preprocessing, noise reduction, feature scaling, and classification using Support Vector Machine (SVM) and Logistic Regression. The system performs binary classification of MRI scans (tumor vs. non-tumor), with SVM showing better accuracy in comparisons.

The literature review shows that although advanced deep learning models and hybrid systems exist, they are often complex, data-intensive, and less practical in resource-limited settings. Classical machine learning methods remain competitive when combined with proper preprocessing and normalization.

Conclusion

Brain tumor diagnosis is one of the most challenging diagnostics in medical imaging because of the complex anatomical structure of the brain, variability in the appearance of tumors, and the need for early detection and accurate diagnosis that provides better survival rates. All these factors make manual interpretation by radiologists cumbersome and prone to inter-observer variability. This therefore calls for effective automated diagnostic systems. Various approaches for the detection of brain tumors, utilizing machine learning and deep learning, have been put forward, but most of the existing approaches suffer from such issues as high computational complexity, dependence on large annotated datasets, being sensitive to noise, and poor generalization across diverse MRI acquisition protocols. To overcome these obstacles, this research work has proposed a machine learning-based approach named NeuroScan for the detection of brain tumors through MRI images, which is efficient and automated. This approach has the capacity to perform robust image processing, feature normalization, and classification together under a single environment. Because the approach has utilized the classifiers SVM and LR, it is capable of effective discrimination between the tumor and non-tumor images of the MRI, while the computational complexity and speed are also low. The comparison has also identified the supremacy of the SVM classifier over the LR classifier, which can be attributed to the robustness of the former for high- dimensional spaces and non-linear boundaries, as reported by various authors, often observable in images. Experimentations have confirmed the effectiveness of the NeuroScan approach, as it obtains an excellent classification accuracy, precision, recall, and F1-measure. NeuroScan clearly shows the effectiveness of a well-designed preprocessing module, feature normalization module, along with traditional machine learning classifiers, over deep learning-based approaches, which require a lot of computation. This makes NeuroScan a promising tool for its adoption in resource-scarce environments. Future research will be directed at extending the NeuroScan framework to multi-class tumor analysis, including localization and segmentation, and combining different deep models. The use of multimodal MRI information, together with validation studies in real-world healthcare settings, will also be considered. The proposed work will lay the foundations of a comprehensive system that will be able to meet the needs of the healthcare sector.

References

[1] Pereira, S., Pinto, A., Alves, V., & Silva, C. A. (2016). Brain tumor segmentation using convolutional neural networks in MRI images. IEEE Transactions on Medical Imaging, 35(5), 1240–1251. [2] Bauer, S., Wiest, R., Nolte, L. P., & Reyes, M. (2013). A survey of MRI-based medical image analysis for brain tumor studies. Physics in Medicine & Biology, 58(13), R97–R129. [3] Litjens, G., et al. (2017). A survey on deep learning in medical image analysis. Medical Image Analysis, 42, 60–88. [4] Menze, B. H., et al. (2015). The multimodal brain tumor image segmentation benchmark (BRATS). IEEE Transactions on Medical Imaging, 34(10), 1993–2024. [5] Mohsen, H., El-Dahshan, E. S. A., El-Horbaty, E. S. M., & Salem, B. M. (2018). Classification using deep learning neural networks for brain tumors. Future Computing and Informatics Journal, 3(1), 68–71. [6] Abiwinanda, N., et al. (2019). Brain tumor classification using convolutional neural network. World Congress on Medical Physics and Biomedical Engineering, 183–189. [7] Havaei, M., et al. (2017). Brain tumor segmentation with deep neural networks. Medical Image Analysis, 35, 18–31. [8] Chaplot, S., Patnaik, L. M., & Jagannathan, N. R. (2006). Classification of magnetic resonance brain images using wavelets as input to support vector machine and neural network. Biomedical Signal Processing and Control, 1(1), 86–92. [9] Nayak, D. R., Dash, R., & Majhi, B. (2016). Brain MR image classification using two-dimensional discrete wavelet transform and AdaBoost with random forests. Neurocomputing, 177, 188–197. [10] Ortiz, A., et al. (2012). Automated detection of brain tumors based on MRI using machine learning techniques. Pattern Recognition Letters, 33(12), 1495–1504. [11] Islam, J., & Zhang, Y. (2018). Brain MRI analysis for tumor detection using deep convolutional neural networks. Medical & Biological Engineering & Computing, 56, 191–201. [12] Kumar, S., Dabas, C., & Godara, S. (2017). Classification of brain MRI tumor images: A hybrid approach. Procedia Computer Science, 122, 510–517. [13] Zhang, Y., Dong, Z., Wu, L., & Wang, S. (2011). A hybrid method for MRI brain image classification. Expert Systems with Applications, 38(8), 10049–10053. [14] Acharya, U. R., et al. (2019). Automated detection of brain abnormality using MRI images. Journal of Medical Systems, 43(9), 1–11. [15] Tandel, G. S., et al. (2020). A review on a deep learning perspective in brain cancer classification. Cancers, 12(1), 111. [16] Rehman, A., et al. (2020). Brain tumor detection and classification using deep learning. IEEE Access, 8, 130580–130593. [17] Afshar, P., et al. (2019). Brain tumor classification using capsule networks. Medical Image Analysis, 58, 101533.

Copyright

Copyright © 2026 S. Gowtham , S. Hemalatha , A. Reshma , T. Priyadharshini, M. Swetha . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.