A validated custom pipeline for three-dimensional kidney stone renderings tocreate an open access repository

Area of Science:

• Urology
• Medical Imaging
• Computer Science

Background:

• Three-dimensional (3D) rendering of urologic pathology is crucial for education and research.
• A lack of standardized, open-access repositories for high-fidelity kidney stone models stratified by chemical composition exists.

Purpose of the Study:

• To develop and validate a reproducible photogrammetry-based pipeline for generating realistic 3D kidney stone renderings.
• To create an open-access repository of these models.

Main Methods:

• Photogrammetry pipeline using custom rotating stage and dual 4K cameras.
• Photographed chemically characterized human stones (calcium oxalate monohydrate, uric acid, cystine, magnesium ammonium phosphate/carbonate apatite, calcium hydrogen phosphate dihydrate).
• Endourologists evaluated rendered models using a 5-point Likert scale for geometric and surface texture fidelity.

Main Results:

• Successful 3D renderings were obtained for calcium oxalate monohydrate, uric acid, magnesium ammonium phosphate/carbonate apatite, and calcium hydrogen phosphate dihydrate stones.
• Uric acid and calcium oxalate monohydrate stones achieved the highest fidelity scores (mean 3.8-3.9).
• Significant differences in geometric fidelity were observed across stone compositions (p=0.026), with moderate inter-rater reliability for aggregated ratings (ICC=0.67).

Conclusions:

• The validated workflow enables the creation of realistic, open-access 3D kidney stone models.
• This resource supports simulation-based education, surgical training, and computer vision research in endourology.
• The repository is available at github.com/uro-glidar/3d-rendering-diverse-stones.