Brain Tumor Detection System

Abstract

Deep-Learning Applied to early identification of brain tumors: An introductory review of the state-of-the-art(s) in under 30 minutes. Brain tumors represent an increasing public point-health concern with about 250000 new cases all over the globe annually and approximately 150000 deaths every year. Since late diagnosis significantly reduces the chance of survival, medical experts are resorting to computerized image analysis to enhance rapidity and clarity in detection of results of MRI surveys. Demarcation of tumors which is done by hand is not only time- consuming but also shows high variability among experts. Third, small lesions, varying contrast and tortuous brain structure, which appear in conventional segmentation found in the tradi- tional techniques of segmenting the brain, such as thresholding and watershed and the use of wavelet transforms, makes the brain hard to segment. The convolutional networks become contempo- rary and can learn tumor-specific patterns with direct learning with data: To-date models to proposal: Accurate irregularity RPN A Faster-CNN one that initially proposes candidate regions (RPN) which are then classified as glioma, meningioma, or pituitary. This approach has been tested on a publicly available MRI dataset and achieved the results shown below; The average of the responses of the detection and classification processes show that the algorithms have a precision of 75.18 percent with glioma, 89.45 percent with meningioma, and 68.18 percent with pituitary tumors. The total mean average precision (mAP) of the model achieved was 77.60 percent with all tumor types.

Introduction

The text focuses on the growing importance of automated brain tumor detection and segmentation using deep learning, particularly YOLO-based models. Brain tumors are a serious global health issue with increasing incidence and high mortality rates, especially for aggressive types like glioblastoma. MRI and CT scans are key diagnostic tools, but manual analysis is time-consuming and prone to human error, creating a need for faster and more accurate AI-based solutions.

Recent research shows that deep learning models such as YOLO and Faster R-CNN significantly improve tumor detection and segmentation performance. While earlier models achieved high accuracy, they struggled with small tumor detection and scalability issues. Newer hybrid and improved architectures aim to enhance feature extraction, localization, and real-time processing, with YOLOv11 using advanced components like CSPDarkNet, SPPF, and anchor-free detection heads to improve accuracy and efficiency.

The proposed system focuses on real-time brain tumor detection and classification (glioma, meningioma, and pituitary tumors) using YOLOv11. It includes preprocessing steps such as resizing MRI images, normalization, and data augmentation to improve robustness. The dataset (from Roboflow Universe) is expanded through augmentation techniques like flipping and rotation to improve generalization. The system is designed to handle variations in image quality and ensure reliable performance across different clinical conditions.

Conclusion

This research result validates the potential of deep learning algorithms to predict the presence of brain tumors with high precision and sensitivity in case of MRI scans. The system is able to localize with high precision and in real-time a variety of MRI conditions and forms of tumors using the YOLOv11 model. Also, preprocessing and augmentation of the data is effective to increase the accuracy and the generalization of the system to be used in a wider range of clinical environments. The YOLOv11-based detection system is faster and more accurate than previous versions, so it is more appropriate to use in the clinical practice. It assists by automatically identifying the location of tumor in order to reduce the workload of radiologist, reduced diagnostic error, and early intervention due to which brain tumor patients can have a better chance to survive. The system can be extended to scan 3D MRI images in the future, segmentation models to facilitate a better delineation of the boundaries, and connected to the working system of hospitals to aid in real-time diagnosis. The study eventually helps in the development of effective and dependable AI-driven diagnosis instruments to the clinical practice.

References

[1] N. M. Dipu, S. A. Shohan and K. M. A Salam, ”Brain Tumor De- tection Using Various Deep Learning Algorithms,” 2021 International Conference on Science Contemporary Technologies (ICSCT), Dhaka, Bangladesh, 2021. [2] A. M. Qarni, S. Anwar and G. M. Khan, ”YOLOv11-Based Brain Tumor Detection and Segmentation for Clinical Decision Support,” 2025 International Conference on Emerging Technologies in Electronics, Computing, and Communication (ICETECC), Jamshoro, Pakistan, 2025. [3] Y. Bhanothu, A. Kamalakannan and G. Rajamanickam, ”Detection and Classification of Brain Tumor in MRI Images using Deep Convolutional Network,” 2020 6th International Conference on Advanced Computing and Communication Systems (ICACCS), Coimbatore, India, 2020. [4] M. V. Madhu Thota and B. R. Prasad, ”INCS: A Novel Brain Tumor Detection Methodology Design Based on Deep Learning Assisted Intel- ligent Neural Classification Strategy,” 2024 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES), Chennai, India, 2024, [5] S. Koshti, V. Degaonkar, I. Modi, I. Srivastava, J. Panambor and A. Jagtap, ”Brain Tumor Detection System using Convolutional Neural Network,” 2022 IEEE Pune Section International Conference (PuneCon), Pune, India, 2022. [6] A. Goswami and M. Dixit, ”An Analysis of Image Segmentation Methods for Brain Tumor Detection on MRI Images,” 2020 IEEE 9th International Conference on Communication Systems and Network Technologies (CSNT), Gwalior, India, 2020. [7] A. E. Babu, A. Subhash, D. S. Rajan, F. Jacob, and P. A. Kumar, “A Survey on Methods for Brain Tumor Detection,” International Journal of Engineering Research Technology (IJERT), vol. 10, no. 4, pp. 1– 5, 2021. [8] K. Agrawal, K. (“Kushagra”), and L. Goel, “Brain Tumor Detection us- ing Machine Learning,” International Journal of Computer Applications, vol. 185, no. 9, May 2023. [9] Ali, A.,Nadeem, M. B., Aziz, M. W., Ashraf, M. W., Mustafa, G. (2025). YOLO-v9-YOLO-v11: Brain Tumor Performance Analysis Using MRI Images. The Asian Bulletin of Big Data Management, 5(3), 135-153. [10] Chourib, I. (2025). From detection to diagnosis: An advanced transfer learning pipeline using YOLO11 with morphological post-processing for brain tumor analysis for MRI images. Journal of Imaging, 11(8), 282. [11] Abuhammad, H. Z., ALsaidi, R. A., Alraqad, R. H. (2025, April). Enhancing Brain Tumor Detection with YOLOv11: An Innovative Deep Learning Approach in Medical Imaging. In 2025 1st International Conference on Computational Intelligence Approaches and Applications (ICCIAA) (pp. 1-6). IEEE. [12] Wahidin, M. F., Kosala, G. (2025, February). Brain Tumor Detection Using YOLO Models in MRI Images.In 2025 International Conference on Advancement in Data Science, E-learning and Information System (ICADEIS) (pp. 1-6). IEEE. [13] Thakur, M. K., Chauhan, S. (2025, April). Brain Tumor Detection and Classification Using YOLO Algorithms. In 2025 International Conference on Inventive Computation Technologies (ICICT) (pp. 116- 123). IEEE. [14] Jeyaraj M, P., Kumar M, S. (2025). Automated Brain Tumor Segmen- tation using Hybrid YOLO and SAM. Current Medical Imaging, 21(1), E15734056392711. [15] Yang, T., Lu, X., Yang, L., Yang, M., Chen, J., Zhao, H. (2025). Application of MRI image segmentation algorithm for brain tumors based on improved YOLO. Frontiers in Neuroscience, 18, 1510175. [16] Hechkel, W.,Missaoui, R., Helali, A. (2025, February). Early Diagnosis of Alzheimer’s Disease Using YOLOv11 Deep Learning Model in MRI Scans. In 2025 IEEE 22nd International Multi-Conference on Systems, Signals Devices (SSD) (pp. 599-604). IEEE. [17] Elgohr, A.T., Elhadidy, M. S., El-geneedy, M., Akram, S., Mousa, M. A. (2025, August). Advancing sign language recognition: a YOLO v. 11-Based deep learning framework for alphabet and transactional hand gesture detection. In Proceedings of the AAAI Symposium Series (Vol. 6, No. 1, pp. 209-217). [18] Behzadi, S., Sharifrazi, D., Mesbahzadeh, B., Joloudari, J. H., Al- izadehsani, R. (2025). A Lightweight and Robust Framework for Real- Time Colorectal Polyp Detection Using LOF-Based Preprocessing and YOLO-v11n. arXiv preprint arXiv:2507.10864.

Copyright

Copyright © 2026 Arpit Tyagi, Anjali Tomar, Arin Lamba. 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.