This project uses the BRATS(Brain Tumor Segmentation Challenge) 2018 dataset to grade different types of brain tumors (HGG and LGG) by extracting features from the different MRI modalities present in the dataset.
The dataset contains 4 3D MRI modalities: T1-weighted, T2-weighted, T1-contrast enhanced, FLair images. The segmented image(images which have only the tumor part) has 3 Region of Interest labeled as 1,2,4. I have made 3 segmented images from the dataset as follows:
- Using only the region labeled '1'.
- Using region labeled '1' and '4'.
- Using region labeled '1', '2' and '4'.
Now, the other 4 images (t1,t2,t1c and flair) are fused using DWT(discrete wavelet transform). The fused image and the 3 segmented images are used to extract features by Pyradiomics and different models are applied to get the results.
The feature selection techniques used on the data are:
- MRMR
- Relief
- Select K-Best
- PCA
The models used are Logistic Regression, Random forest and XG Boost to get the results. The highest accuracy achieved is 92.95%.
If you find any problem with the repository, please report an issue or if you have to give any feedback, please mail me at [email protected]
[1] B. H. Menze, A. Jakab, S. Bauer, J. Kalpathy-Cramer, K. Farahani, J. Kirby, et al. "The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)", IEEE Transactions on Medical Imaging 34(10), 1993-2024 (2015) DOI: 10.1109/TMI.2014.2377694
[2] S. Bakas, H. Akbari, A. Sotiras, M. Bilello, M. Rozycki, J.S. Kirby, et al., "Advancing The Cancer Genome Atlas glioma MRI collections with expert segmentation labels and radiomic features", Nature Scientific Data, 4:170117 (2017) DOI: 10.1038/sdata.2017.117
[3] S. Bakas, M. Reyes, A. Jakab, S. Bauer, M. Rempfler, A. Crimi, et al., "Identifying the Best Machine Learning Algorithms for Brain Tumor Segmentation, Progression Assessment, and Overall Survival Prediction in the BRATS Challenge", arXiv preprint arXiv:1811.02629 (2018)
[4] van Griethuysen, J. J. M., Fedorov, A., Parmar, C., Hosny, A., Aucoin, N., Narayan, V., Beets-Tan, R. G. H., Fillon-Robin, J. C., Pieper, S., Aerts, H. J. W. L. (2017). Computational Radiomics System to Decode the Radiographic Phenotype. Cancer Research, 77(21), e104–e107. https://doi.org/10.1158/0008-5472.CAN-17-0339 <https://doi.org/10.1158/0008-5472.CAN-17-0339>_
[5] Hanchuan Peng, Fuhui Long, and Chris Ding, "Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 27, No. 8, pp.1226-1238, 2005.