README.md

About Brain MRI Dataset

• This brain tumor dataset containing 3064 T1-weighted contrast-enhanced MRI (T1w CE-MRI) images from 233 patients with three kinds of brain tumor:

Sr. No.	Tumor Name	Number of Observations
1	Meningioma	708
2	Glioma	1426
3	Pituitary Tumor	930

• This dataset is organized in MATLAB data (.mat file) format which is one of the data-exchange format by MATLAB. Each .mat file stores a struct containing the following fields for an MRI image:

cjdata.label: 1 for meningioma, 2 for glioma, 3 for pituitary tumor
cjdata.PID: patient ID
cjdata.image: image data
cjdata.tumorBorder: a vector storing the coordinates of discrete points on tumor border.
		For example, [x1, y1, x2, y2,...] in which x1, y1 are planar coordinates on tumor border.
		It was generated by manually delineating the tumor border. So we can use it to generate
		binary image of tumor mask.
cjdata.tumorMask: a binary image with 1s indicating tumor region

• The jupyter notebook preprocessing_mat_files.ipynb is also provided which consist of python code to extract the information using h5py library which provide a File class to open and process .mat files.
• The following function is written to process the .mat file and extract the different image related information which is stored in an instance of a File class of h5py library with a dictionary kind of format. The rest of the source-code exports the images into the required directories (created on the basis of tumor classes) after image data is being extracted from the .mat file using the following function:

def mat_file_to_dict(filepath: str) -> dict:
	tumor_class = {1: 'meningioma', 2: 'glioma', 3: 'pituitary_tumor'}
	tumor_data_dict = {}
	with h5py.File(filepath, mode = 'r') as image_data:
		cjdata_struct = image_data['cjdata']
		tumor_data_dict['class'] = tumor_class[int(cjdata_struct['label'][0, 0])]
		tumor_data_dict['image'] = cjdata_struct['image'][:].transpose()
		tumor_data_dict['tumor_border'] = cjdata_struct['tumorBorder'][0]
		tumor_data_dict['tumor_mask'] = cjdata_struct['tumorMask'][:].transpose()
	return tumor_data_dict

• The dataset was used in the following research papers:
  1. Cheng, Jun, et al. "Enhanced Performance of Brain Tumor Classification via Tumor Region Augmentation and Partition." PloS one 10.10 (2015).
  2. Cheng, Jun, et al. "Retrieval of Brain Tumors by Adaptive Spatial Pooling and Fisher Vector Representation." PloS one 11.6 (2016).

NOTE: MATLAB source-codes are available on the repository: https://github.com/chengjun583/brainTumorRetrieval