Published August 2, 2021 | Version v1.0.0
Dataset Open

T2-weighted Kidney MRI Segmentation

Description

A dataset containing 100 T2-weighted abdominal MRI scans and manually defined kidney masks. This MRI sequence is designed to optimise contrast between the kidneys and surrounding tissue to increase the accuracy of segmentation. Half of the acquisitions were acquired of healthy control subjects while the other half were acquired from Chronic Kidney Disease (CKD) patients. Ten of the subjects were scanned five times in the same session to enable assessment of the precision of Total Kidney Volume (TKV) measurements. More information about each subject can be found in the included csv file. This dataset was used to train a Convolutional Neural Network (CNN) to automatically segment the kidneys. 

For more information about the dataset please refer to this article.

For an executable that allows automated segmentation of the kidneys from this dataset please refer to this software.

Files

CKD.zip

Files (160.3 MB)

Name Size Download all
md5:8fe112954083212cbee80a6087db84c5
83.3 MB Preview Download
md5:561badbc875874c5e69cf725a8545330
77.0 MB Preview Download
md5:3bc3d595339cb7c1f9208a07f7335ef5
1.2 kB Preview Download

Additional details

Related works

Is documented by
Journal article: 10.1002/mrm.28768 (DOI)
Is source of
Software: 10.5281/zenodo.4068851 (DOI)

Funding

UK Research and Innovation
EPSRC and MRC Centre for Doctoral Training in Biomedical Imaging EP/L016052/1
UK Research and Innovation
UK Renal Imaging Network (UKRIN): Enabling clinical translation of functional MRI for kidney disease MR/R02264X/1

References

  • Daniel AJ, Buchanan CE, Allcock T, et al. Automated renal segmentation in healthy and chronic kidney disease subjects using a convolutional neural network. Magnetic Resonance in Medicine 2021;86:1125–1136 doi: https://doi.org/10.1002/mrm.28768.