Glioma brain tumour classification and detection of magnetic resonance imaging using convolutional neural network / Syazani Adriana Abdul Ghaffar
In Malaysia, brain tumour is uncommon cancer in comparison to other types of cancer, only 1.95% of malignancies cases have been reported by the Malaysian Journal of Public Health Medicine (2017). There are numerous types of brain tumours have been identified, such as Glioma. This glioma abnormality...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 語言: | English |
| 出版: |
2023
|
| 主題: | |
| 在線閱讀: | https://ir.uitm.edu.my/id/eprint/89007/1/89007.pdf |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | In Malaysia, brain tumour is uncommon cancer in comparison to other types of cancer, only 1.95% of malignancies cases have been reported by the Malaysian Journal of Public Health Medicine (2017). There are numerous types of brain tumours have been identified, such as Glioma. This glioma abnormality occurs in the brain and spinal cord and is one of the most common types of primary brain tumours. It is the most aggressive and fatal type of tumour. Magnetic Resonance Imaging is an effective noninvasive method to detect presence of brain tumour, but it has limitations. The problem that is addressed in this research is that the manual evaluation of detecting brain tumours consumes time and to able to classify brain tumour, feature extraction needed to be done but it is complex. Besides that, noise interference may affect tumour classification accuracy. This research uses Convolution Neural Network to classify and detect the MRI brain image. Hence the objective of this research is to design and develop a prototype of glioma brain tumour classification and detection of MRI brain images using CNN. Lastly, evaluation is done to test the accuracy, functionality, and usability of the proposed prototype and had achieved 99.00% accuracy, 100.00% precision, 98.00% recall. The proposed method of detection on MRI brain images accurately classifies and detect the image and achieving a great score of classification accuracy. With further extensive research, the system can be improved with detecting more classes of MRI brain images and able to detect the location of the abnormalities in the brain region. |
|---|