BRAIN TUMOUR CLASSIFICATON AND SEGMENTATION USING DEEP LEARNING

  • Unique Paper ID: 170305
  • Volume: 11
  • Issue: 7
  • PageNo: 1300-1312
  • Abstract:
  • Among all types of cancers, brain cancer has been among the most common affecting many people. The disease is now endangering lives. Early detection is crucial for making life-saving interventions. MRI is a very powerful device for detecting different brain abnormalities and widely used by radiologists and physicians. We propose Deep learning-based convolutional neural network techniques to identify different types of brain cancers. The proposed model uses large datasets with five classes (meningiomas, gliomas, pituitary, Neurocitoma, Schwannoma tumors). The model is developed based on Residual Network or ResNet-50 and has a special designed architecture to learn abnormal complex data features. This helps doctors decide whether the patient has the disease. With deep learning technology, it offers a more accurate, effective, faster way to identify brain tumors. The proposed method achieves accuracy of 99.08%. Experimental results show the efficiency of the proposed method for BT multi-class categorization. A brain tumor is when there is an abnormal growth of cells somewhere in the brain or in its central nervous system. It can be primary or it can metastasize. A Magnetic Resonance Imaging scan (MRI) is one of the most common means of detection of brain tumors. However, the modern-day best-in-class format for brain tumor segmentation such as the U-Net architecture would still be preferred because radiologists may differ in perspective. In America only, over 160.000 people live with brain and other nervous system cancers, and thus timely detection of these tumors becomes vital. In this said paper, three models; Regular U-Net, Upgraded U-net, and Attention U-net are shown to segmentation against the brain tumor and compared using Dice Score Coefficient (DSC). Among these three models, regular U-Net renders the best DSC of 0.3902, 0.6877, and 0.6534 for the Necrotic, Edema, and Enhancing Tumor mask, respectively. These include promising findings as well as floors for future improvements. Research might investigate new methods of pre-processing data or alternate designs of model architecture to bring about an improvement in segmentation accuracy. Such findings illustrate the potential of advanced deep learning models like UNet in brain tumor detection, along with adding and comparing various models and techniques to further enhance each model's performance for more coherent and accurate results critical in timely and right patient care.

Copyright & License

Copyright © 2025 Authors retain the copyright of this article. This article 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.

BibTeX

@article{170305,
        author = {Sourav Mandot Alais Jain and Harsh Jadhav and Pranjal Pandit},
        title = {BRAIN TUMOUR CLASSIFICATON AND SEGMENTATION USING DEEP LEARNING},
        journal = {International Journal of Innovative Research in Technology},
        year = {2024},
        volume = {11},
        number = {7},
        pages = {1300-1312},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=170305},
        abstract = {Among all types of cancers, brain cancer has been among the most common affecting many people. The disease is now endangering lives. Early detection is crucial for making life-saving interventions. MRI is a very powerful device for detecting different brain abnormalities and widely used by radiologists and physicians. We propose Deep learning-based convolutional neural network techniques to identify different types of brain cancers. The proposed model uses large datasets with five classes (meningiomas, gliomas, pituitary, Neurocitoma, Schwannoma tumors). The model is developed based on Residual Network or ResNet-50 and has a special designed architecture to learn abnormal complex data features. This helps doctors decide whether the patient has the disease. With deep learning technology, it offers a more accurate, effective, faster way to identify brain tumors. The proposed method achieves accuracy of 99.08%. Experimental results show the efficiency of the proposed method for BT multi-class categorization.
A brain tumor is when there is an abnormal growth of cells somewhere in the brain or in its central nervous system. It can be primary or it can metastasize. A Magnetic Resonance Imaging scan (MRI) is one of the most common means of detection of brain tumors. However, the modern-day best-in-class format for brain tumor segmentation such as the U-Net architecture would still be preferred because radiologists may differ in perspective. In America only, over 160.000 people live with brain and other nervous system cancers, and thus timely detection of these tumors becomes vital. In this said paper, three models; Regular U-Net, Upgraded U-net, and Attention U-net are shown to segmentation against the brain tumor and compared using Dice Score Coefficient (DSC). Among these three models, regular U-Net renders the best DSC of 0.3902, 0.6877, and 0.6534 for the Necrotic, Edema, and Enhancing Tumor mask, respectively. These include promising findings as well as floors for future improvements. Research might investigate new methods of pre-processing data or alternate designs of model architecture to bring about an improvement in segmentation accuracy. Such findings illustrate the potential of advanced deep learning models like UNet in brain tumor detection, along with adding and comparing various models and techniques to further enhance each model's performance for more coherent and accurate results critical in timely and right patient care.},
        keywords = {MRI, U-net, ResNet50},
        month = {December},
        }

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 7
  • PageNo: 1300-1312

BRAIN TUMOUR CLASSIFICATON AND SEGMENTATION USING DEEP LEARNING

Related Articles