Preprocessed IXI Dataset
IXI Dataset Description
IXI Dataset contains nearly 600 MR images from normal, healthy subjects.
- T1, T2, and PD-weighted images
- MRA images
- Diffusion-weighted images (15 directions)
- Hammersmith Hospital using a Philips 3T system
- Guy’s Hospital using a Philips 1.5T system
- Institute of Psychiatry using a GE 1.5T system
The IXI dataset is available under the Creative Commons CC BY-SA 3.0 license. For more detailed information and download, please refer to the official website of IXI project
The T1 MRI data were used and preprocessed to generate normalised brain volume maps. The grey matter (GM) and white matter (WM) images were analyzed together.
- AC-PC Realignment
- GM, WM Tissue Segmentation
- Non-linear registration to MNI152 space
- 4mm Smoothing
The result, normalized brain volume map, has shape of 121x145x121. Each voxel in the volume map represent regional volume of tissue.
All images were pre-processed using SPM12 in the MATLAB R2018a environment. DARTEL which is one of the SPM extension was used for normalization, non-linear registeration, resampling, modulation and smoothing step.
We followed the protocol from the paper "Predicting brain age with deep learning from raw imaging data results in a reliable and heritable biomarker" by James H. Cole et al. For the very detailed preprocessing tutorial of Voxel-based Morphometry(VBM), see here
Before we upload the preprocessed data to the Server of XAI Center, we can access to the dataset through this temporary link
A machine learning and statistical inference framework for explainable artificial intelligence(의사결정 이유를 설명할 수 있는 인간 수준의 학습·추론 프레임워크 개발)
Ministry of Science and ICT/XAIC
UNIST, Korea Univ., Yonsei Univ., KAIST., AItrics
This data is made available under the Creative Commons CC BY-SA 3.0 license.