Knee osteoarthritis (OA) is a painful joint disease, causing disabilities in daily activities. However, there is no known cure for OA, and the best treatment strategy might be prevention. Finite element (FE) modeling has demonstrated potential for evaluating personalized risks for the progression of OA. Current FE modeling approaches use primarily magnetic resonance imaging (MRI) to construct personalized knee joint models. However, MRI is expensive and has lower resolution than computed tomography (CT). In this study, we extend a previously presented atlas-based FE modeling framework for automatic model generation and simulation of knee joint tissue responses using contrast agent-free CT. In this method, based on certain anatomical dimensions measured from bone surfaces, an optimal template is selected and scaled to generate a personalized FE model. We compared the simulated tissue responses of the CT-based models with those of the MRI-based models. We show that the CT-based models are capable of producing similar tensile stresses, fibril strains, and fluid pressures of knee joint cartilage compared to those of the MRI-based models. This study provides a new methodology for the analysis of knee joint and cartilage mechanics based on measurement of bone dimensions from native CT scans.

膝关节骨关节炎 (OA) 是一种疼痛的关节疾病，会导致日常活动障碍。然而，OA 没有已知的治愈方法，最好的治疗策略可能是预防。有限元 (FE) 建模已显示出评估 OA 进展的个性化风险的潜力。当前的有限元建模方法主要使用磁共振成像 (MRI) 来构建个性化的膝关节模型。然而，MRI 价格昂贵且分辨率低于计算机断层扫描 (CT)。在这项研究中，我们扩展了先前提出的基于图谱的 FE 建模框架，用于使用无造影剂 CT 自动生成模型和模拟膝关节组织反应。在这种方法中，基于从骨表面测量的某些解剖尺寸，选择并缩放最佳模板以生成个性化的有限元模型。我们将基于 CT 的模型的模拟组织反应与基于 MRI 的模型的模拟组织反应进行了比较。我们表明，与基于 MRI 的模型相比，基于 CT 的模型能够产生类似的拉伸应力、原纤维应变和膝关节软骨的流体压力。这项研究提供了一种新的方法来分析膝关节和软骨力学，该方法基于对原生 CT 扫描的骨骼尺寸的测量。