Objective

To assess the ability of imaging-based deep learning to detect radiographic patellofemoral osteoarthritis (PFOA) from knee lateral view radiographs.

Design

Knee lateral view radiographs were extracted from The Multicenter Osteoarthritis Study (MOST) public use datasets (n = 18,436 knees). Patellar region-of-interest (ROI) was first automatically detected, and subsequently, end-to-end deep convolutional neural networks (CNNs) were trained and validated to detect the status of patellofemoral OA. Patellar ROI was detected using deep-learning-based object detection method. Atlas-guided visual assessment of PFOA status by expert readers provided in the MOST public use datasets was used as a classification outcome for the models. Performance of classification models was assessed using the area under the receiver operating characteristic curve (ROC AUC) and the average precision (AP) obtained from the Precision-Recall (PR) curve in the stratified 5-fold cross validation setting.

Results

Of the 18,436 knees, 3,425 (19%) had PFOA. AUC and AP for the reference model including age, sex, body mass index (BMI), the total Western Ontario and McMaster Universities Arthritis Index (WOMAC) score, and tibiofemoral Kellgren–Lawrence (KL) grade to detect PFOA were 0.806 and 0.478, respectively. The CNN model that used only image data significantly improved the classifier performance (ROC AUC = 0.958, AP = 0.862).

Conclusion

We present the first machine learning based automatic PFOA detection method. Furthermore, our deep learning based model trained on patella region from knee lateral view radiographs performs better at detecting PFOA than models based on patient characteristics and clinical assessments.

中文翻译：

---

使用深度学习从膝关节侧位片中自动检测髌股骨关节炎：来自多中心骨关节炎研究 (MOST) 的数据

客观的

评估基于成像的深度学习从膝关节侧位片中检测 X 线髌股骨关节炎 (PFOA) 的能力。

设计

从多中心骨关节炎研究 (MOST) 公共使用数据集 ( n = 18,436 膝）。首先自动检测髌骨感兴趣区域 (ROI)，随后训练和验证端到端深度卷积神经网络 (CNN) 以检测髌股骨 OA 的状态。使用基于深度学习的对象检测方法检测髌骨 ROI。大多数公共使用数据集中提供的专家读者对全氟辛酸状态的图集引导视觉评估被用作模型的分类结果。使用受试者工作特征曲线下面积 (ROC AUC) 和从分层 5 折交叉验证设置中的精确召回 (PR) 曲线获得的平均精确度 (AP) 评估分类模型的性能。

结果

在 18,436 个膝盖中，3,425 个（19%）有 PFOA。参考模型的 AUC 和 AP 包括年龄、性别、体重指数 (BMI)、西安大略大学和麦克马斯特大学关节炎指数 (WOMAC) 总评分以及检测 PFOA 的胫股 Kellgren-Lawrence (KL) 等级分别为 0.806 和 0.478，分别。仅使用图像数据的 CNN 模型显着提高了分类器性能（ROC AUC = 0.958，AP = 0.862）。

结论

我们提出了第一个基于机器学习的自动 PFOA 检测方法。此外，与基于患者特征和临床评估的模型相比，我们在膝盖侧位 X 光片上对髌骨区域进行训练的基于深度学习的模型在检测 PFOA 方面表现更好。