/brain-mri-modality-classification

MRI modality(T1, T2, FLAIR) classification model with modified ResNet-50. Hanyang univ. dep. of biomedical engineering graduation project.

Primary LanguagePython

Brain MRI Modality(T1, T2, FLAIR) Classification with Modified ResNet-50

Usage

  1. Preprocess your MRI images(3D NIfTI) with python scripts/preprocess_images.py
  2. Train your model with python scripts/train.py
  3. Test your model with python scripts/test.py default oasis3

Datasets

Dataset Modality Details
ADNI1 T1 MPRAGE
ADNI1 T2 T2-FSE
ADNI2 T1 MPRAGE
ADNI2 FLAIR FLAIR (Axial)
ADNI3 T1 MPRAGE (Sagittal)
ADNI3 FLAIR FLAIR (Sagittal)
ADNIGO T1 MPRAGE
ADNIGO FLAIR FLAIR (Axial)
CAMCAN T1
CAMCAN T2
IXI T1
IXI T2
Kirby-21 T1 MPRAGE
Kirby-21 T2
Kirby-21 FLAIR
MICCAI2017 T1
MICCAI2017 FLAIR
MICCAI2018 T1
MICCAI2018 T2
MICCAI2018 FLAIR

References

  • Çinar, A., & Yildirim, M. (2020). Detection of tumors on brain MRI images using the hybrid convolutional neural network architecture. Medical hypotheses, 139, 109684.