In this study, we propose a novel framework that utilizes deep learning (DL) and attention mechanisms to predict the radiographic progression of patellofemoral osteoarthritis (PFOA) over a period of seven years. This study included subjects (1832 subjects, 3276 knees) from the baseline of the MOST study. PF joint regions-of-interest were identified using an automated landmark detection tool (BoneFinder) on lateral knee X-rays. An end-to-end DL method was developed for predicting PFOA progression based on imaging data in a 5-fold cross-validation setting. A set of baselines based on known risk factors were developed and analyzed using gradient boosting machine (GBM). Risk factors included age, sex, BMI and WOMAC score, and the radiographic osteoarthritis stage of the tibiofemoral joint (KL score). Finally, we trained an ensemble model using both imaging and clinical data. Among the individual models, the performance of our deep convolutional neural network attention model achieved the best performance with an AUC of 0.856 and AP of 0.431; slightly outperforming the deep learning approach without attention (AUC=0.832, AP= 0.4) and the best performing reference GBM model (AUC=0.767, AP= 0.334). The inclusion of imaging data and clinical variables in an ensemble model allowed statistically more powerful prediction of PFOA progression (AUC = 0.865, AP=0.447), although the clinical significance of this minor performance gain remains unknown. This study demonstrated the potential of machine learning models to predict the progression of PFOA using imaging and clinical variables. These models could be used to identify patients who are at high risk of progression and prioritize them for new treatments. However, even though the accuracy of the models were excellent in this study using the MOST dataset, they should be still validated using external patient cohorts in the future.

翻译：本研究提出一种利用深度学习（DL）与注意力机制预测髌股关节骨关节炎（PFOA）七年间影像学进展的新框架。研究纳入MOST研究基线期的受试者（1832名受试者，3276个膝关节），通过自动标志点检测工具（BoneFinder）在侧位膝关节X光片上识别髌股关节感兴趣区域。我们开发了基于影像数据的端到端深度学习预测方法，采用5折交叉验证进行评估。基于已知风险因素建立多组基线模型，采用梯度提升机（GBM）进行分析——风险因素包括年龄、性别、BMI、WOMAC评分及胫股关节影像学骨关节炎分级（KL分级）。最终训练了融合影像与临床数据的集成模型。在单一模型中，深度卷积神经网络注意力模型性能最优（AUC=0.856，AP=0.431），略优于无注意力机制的深度学习方法（AUC=0.832，AP=0.4）及最佳参考GBM模型（AUC=0.767，AP=0.334）。将影像数据与临床变量纳入集成模型后，对PFOA进展的预测能力在统计学上显著提升（AUC=0.865，AP=0.447），但该微小性能增益的临床意义尚不明确。本研究证明机器学习模型利用影像与临床变量预测PFOA进展的潜力，这类模型可用于识别高进展风险患者并优先纳入新治疗方案。然而，尽管基于MOST数据集的模型精度优异，未来仍需在外部患者队列中进行验证。