Korean J Transplant.  2022 Nov;36(Supple 1):S160. 10.4285/ATW2022.F-2796.

Development of an AI model for kidney cortex volumetry for donor evaluation

• Jo EA
• Min S
• Jang Y
• Kim YC
• Kim JS
• Han A
• Ha J

• KMID: 2535676
• DOI: http://doi.org/10.4285/ATW2022.F-2796

Abstract

Background
Computer tomography (CT) images can accurately map vasculatures, identify abnormalities and potentially measure split renal function through length or volume of each kidney. The purpose of this study was to develop and validate an automated method to segment and measure kidney cortex volume on contrast-enhanced abdominal CT images of kidney donors.
Methods
The predonation arterial phase CT DICOM images of living kidney donors were downloaded and uploaded to 'OncoStudio' (OncoSoft Inc., Seoul, South Korea), which was used as the AI-based auto-segmentation tool. The AI model within the OncoStudio has a U-Net structure based on a 3D dense block and automatically proceeds to CT site detection and segmentation without clicking by humans. For this study, a total of 82 datasets were used, 70 for training, two for validation, and 10 for independent testing.
Results
The consistency between manually segmented volumes and automatically segmented volumes based on AI was evaluated. The statistics for a total of 20 organs were calculated by combining the left and right cortex of 10 testing datasets. The Dice similarity coefficient (DSC) representing the degree of agreement between 3D volumes was 0.91, and the Hausdorff dis-tance 95% (HD95) representing the lower 95% distance between 3D surface points was 1.52 mm.
Conclusions
An automated method for measuring kidney cortex volume was successfully developed. The auto-segmentation program can be a time saving and promising evaluation tool for donor suitability and split renal function.