Wang X, Shi G, Sivakumar A, Ye T, Sylvester A, Stayman J W, Zbijewski W
Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD USA 21205.
Center for Functional Anatomy and Evolution, Johns Hopkins University, Baltimore, MD USA 21205.
Proc SPIE Int Soc Opt Eng. 2025 Feb;13410. doi: 10.1117/12.3049125. Epub 2025 Apr 2.
We developed a generative model capable of producing synthetic trabecular bone that can be precisely tuned to achieve specific structural characteristics, such as bone volume fraction (BV/TV), trabecular thickness (Tb.Th), and spacing (Tb.Sp).
The generative model is based on Diffusion Transformers (DiT), a latent diffusion approach employing a transformer architecture in the denoising network. To control the microstructure characteristics of the synthetic trabecular bone samples, the model is conditioned on BV/TV, Tb.Th, and Tb.Sp. The training data involved 29898 256×256-pixel Regions of Interest (ROIs) extracted from micro-CT volumes ( voxel size) of 20 femoral bone specimens, paired with trabecular metrics computed within each ROI; the training/validation split was 9:1. For testing, 3499 synthetic bone samples were generated over a wide range of condition (target) microstructure metrics. Results were evaluated in terms of (i) the ability to cover real-world distribution of trabecular structures (coverage), (ii) agreement with target metric values (Pearson Correlation), and (iii) consistency of the metrics across multiple realizations of the DiT model with fixed condition (Coefficient of Variation, CV).
The model achieved good coverage of real-world bone microstructures and visual similarity to true trabecular ROIs. Pearson Correlations against the condition (target) metric values were high: 0.9540 for BV/TV, 0.9618 for Tb.Th, and 0.9835 Tb.Sp. Microstructural characteristics of the synthetic samples were stable across DiT realizations, with CV ranging from 3.37% to 11.78% for BV/TV, 2.27% to 3.22% for Tb.Th, and 2.53% to 5.00% for Tb.Sp.
The proposed generative model is capable of generating realistic digital trabecular bones that can be precisely tuned to achieve specified microstructural characteristics. Possible applications include virtual clinical trials of new skeletal image biomarkers and establishing priors for advanced image reconstruction.
我们开发了一种生成模型，该模型能够生成合成小梁骨，且可以对其进行精确调整以实现特定的结构特征，如骨体积分数（BV/TV）、小梁厚度（Tb.Th）和间距（Tb.Sp）。
该生成模型基于扩散变换器（DiT），这是一种在去噪网络中采用变换器架构的潜在扩散方法。为了控制合成小梁骨样本的微观结构特征，该模型以BV/TV、Tb.Th和Tb.Sp为条件。训练数据包括从20个股骨标本的显微CT体积（体素大小）中提取的29898个256×256像素的感兴趣区域（ROI），并与每个ROI内计算的小梁指标配对；训练/验证分割比例为9:1。为了进行测试，在广泛的条件（目标）微观结构指标范围内生成了3499个合成骨样本。结果根据以下方面进行评估：（i）覆盖小梁结构真实世界分布的能力（覆盖率），（ii）与目标指标值的一致性（皮尔逊相关性），以及（iii）在固定条件下DiT模型的多次实现中指标的一致性（变异系数，CV）。
该模型对真实世界的骨微观结构实现了良好的覆盖，并且与真实小梁ROI具有视觉相似性。与条件（目标）指标值的皮尔逊相关性很高：BV/TV为0.9540，Tb.Th为0.9618，Tb.Sp为0.9835。合成样本的微观结构特征在DiT的多次实现中是稳定的，BV/TV的CV范围为3.37%至11.78%，Tb.Th为2.27%至3.22%，Tb.Sp为2.53%至5.00%。
所提出的生成模型能够生成逼真的数字小梁骨，并且可以对其进行精确调整以实现指定的微观结构特征。可能的应用包括新型骨骼图像生物标志物的虚拟临床试验以及为先进图像重建建立先验知识。
