Bone‐GAN: Generation of virtual bone microstructure of high resolution peripheral quantitative computed tomography
Authors:
Open Access
- 1 June 2023
- journal article
- research article
- Published by Wiley in Medical Physics
- Vol. 50 (11) , 6943-6954
Abstract
Background: Data‐driven development of medical biomarkers of bone requires a large amount of image data but physical measurements are generally too restricted in size and quality to perform a robust training.Purpose: This study aims to provide a reliable in silico method for the generation of realistic bone microstructure with defined microarchitectural properties. Synthetic bone samples may improve training of neural networks and serve for the development of new diagnostic parameters of bone architecture and mineralization.Methods: One hundred‐fifty cadaveric lumbar vertebrae from 48 different male human spines were scanned with a high resolution peripheral quantitative CT. After prepocessing the scans, we extracted 10,795 purely spongeous bone patches, each with a side length of 32 voxels (5 mm) and isotropic voxel size of 164 μm. We trained a volumetric generative adversarial network (GAN) in a progressive manner to create synthetic microstructural bone samples. We then added a style transfer technique to allow the generation of synthetic samples with defined microstructure and gestalt by simultaneously optimizing two entangled loss functions. Reliability testing was performed by comparing real and synthetic bone samples on 10 well‐understood microstructural parameters.Results: The method was able to create synthetic bone samples with visual and quantitative properties that effectively matched with the real samples. The GAN contained a well‐formed latent space allowing to smoothly morph bone samples by their microstructural parameters, visual appearance or both. Optimum performance has been obtained for bone samples with voxel size 32 × 32 × 32, but also samples of size 64 × 64 × 64 could be synthesized.Conclusions: Our two‐step‐approach combines a parameter‐agnostic GAN with a parameter‐specific style transfer technique. It allows to generate an unlimited anonymous database of microstructural bone samples with sufficient realism to be used for the development of new data‐driven methods of bone‐biomarkers. Particularly, the style transfer technique can generate datasets of bone samples with specific conditions to simulate certain bone pathologies.Keywords
Funding Information
- Salesforce (Einstein AI Grant)
- Consejo Nacional de Investigaciones Científicas y Técnicas (PIP2021‐11220200102981CO)
This publication has 30 references indexed in Scilit:
- Human trabecular bone microarchitecture can be assessed independently of density with second generation HR-pQCTBone, 2015
- Generalizing the mean intercept length tensor for gray-level imagesMedical Physics, 2012
- Medical Imaging in the Diagnosis of Osteoporosis and Estimation of the Individual Bone Fracture Risk2011
- The osseous endplates in lumbar vertebrae: Thickness, bone mineral density and their associations with age and disk degenerationBone, 2011
- Trabecular Bone Structure Analysis in the Osteoporotic Spine Using a Clinical In Vivo Setup for 64-Slice MDCT Imaging: Comparison to μCT Imaging and μFE ModelingJournal of Bone and Mineral Research, 2009
- Image quality assessment: from error visibility to structural similarityIEEE Transactions on Image Processing, 2004
- A new method for the model‐independent assessment of thickness in three‐dimensional imagesJournal of Microscopy, 1997
- Quantification of Bone Microarchitecture with the Structure Model IndexComputer Methods in Biomechanics and Biomedical Engineering, 1997
- Lumbar Spinal Pathology in Cadaveric Material in Relation to History of Back Pain, Occupation, and Physical LoadingSpine, 1990
- The quantitative morphology of anisotropic trabecular boneJournal of Microscopy, 1974