Superiority of Multiple-Joint Space Width over Minimum-Joint Space Width Approach in the Machine Learning for Radiographic Severity and Knee Osteoarthritis Progression

Abstract

We compared the prediction efficiency of the multiple-joint space width (JSW) and the minimum-JSW on knee osteoarthritis (KOA) severity and progression by using a deep learning approach. A convolutional neural network (CNN) with ResU-Net architecture was developed for knee X-ray imaging segmentation and has attained a segmentation efficiency of 98.9% intersection over union (IoU) on the distal femur and proximal tibia. Later, by leveraging the image segmentation, the minimum and multiple-JSWs in the tibiofemoral joint were estimated and then validated by radiologist measurements in the Osteoarthritis Initiative (OAI) dataset using Pearson correlation and Bland–Altman plots. The agreement between the CNN-based estimation and radiologist’s measurement of minimum-JSWs reached 0.7801 (p < 0.0001). The estimated JSWs were deployed to predict the radiographic severity and progression of KOA defined by Kellgren-Lawrence (KL) grades using the XGBoost model. The 64-point multiple-JSWs achieved the best performance in predicting KOA progression within 48 months, with the area-under-receiver operating characteristic curve (AUC) of 0.621, outperforming the commonly used minimum-JSW with 0.554 AUC. We provided a fully automated radiographic assessment tool for KOA with comparable performance to the radiologists and showed that the fine-grained measurement of multiple-JSWs yields superior prediction performance for KOA over the minimum-JSW.

Discovering Knee Osteoarthritis Imaging Features for Diagnosis and Prognosis: Review of Manual Imaging Grading and Machine Learning Approaches

“…Cheung et al [49] have tested the segmentation ability of four models, namely CUMed-Vision, U-Net, DeepLabv3, and Res-U-Net. All four models were used to segment distal (ii) Detected patellofemoral joints on X-ray images [44] (ii) Quantification of qualitative OA features (ii) Local binary pattern [41,42,44] (iii) Random forest regression voting [44] (iv) Fully convolutional neural network [45,46] (iii) Detected cartilage X-ray images [41,42] (v) YOLOv2 network [46] Segmentation of knee joint components Diagnosis (i) Segmented knee cartilage from 2D ultrasound images [27,28] (i) Area measurement (i) Locally statistical level set method [28] (ii) Segmented knee cartilage from 2D MRI images [47] (ii) Volumetric measurement (ii) Automatic seed point detection [48] (iii) Segmented cartilage and meniscus from MRI images [29] (iii) Joint shape measurement (iii) Random walker [27] (iv) Segmented subchondral bone from multiple 2D MRI images [48] (iv) Quantification of measurable OA features (iv) Watershed (v) Segmented distal femur and proximal tibia from X-ray images [49] (v) Reconstruction of 3D knee joint model for simulation and joint loading study (v) Graph cut [27] (vi) Calculated joint space width on X-ray images [49] (vi) Finite element analysis (vi) Support vector machine classifier [43] (vii) Segmented femoral condyle cartilage from ultrasound images [50] (vii) Utilization of statistical and computational models (vii) Decision tree classifier [41,42] (viii) Segmented bones (femur and tibia) and cartilages (femoral and tibial cartilages) on MRI images [51] (viii) Active contour algorithm [42] (ix) Segmented knee bones, cartilage, and muscle tissues on MRI images [52,53] (ix) U-Net [29,47,…”

“…Prognosis (i) Estimated future knee OA incidence (i) Risk stratification (i) Random forest classifier [64,65] (a) 30 months [64] (ii) Selection of data from suitable time points to indicate short-term and long-term OA changes (ii) Logistic regression classifier [66][67][68][69] (b) 48 months [70] (iii) OA feature change detection (iii) Support vector machine classifier [66] (c) 8 years (iv) Discovery of pain-associated imaging features (iv) XGBoost model [49] (ii) Predicted medial JSN progression [66] (v) Multilayer perceptron [67,71] (iii) Predicted radiographic joint space loss progression [67] (vi) LASSO regression [39] (iv) Predicted knee OA onset and knee OA deterioration [71] (vii) Artificial neural network [70] (v) Discriminated between progressors and nonprogressors [72] (viii) Deep CNN [44,72,73] (vi) Predicted pain [73,74] (ix) DenseNet CNN [68] (vii) Predicted risk of progressive pain and structural change [65] (x) Gradient boosting machine [44,70] (viii) Predicted total knee replacement (TKR) incidence [68,75] (xi) Duo classifier [65] (xii) DeepSurv [75] (xiii) Dynamic functional mixedeffects model [54] femur and proximal tibia. Res-U-Net gave the best segmentation outcome with the highest mean intersection over union score at 0.989.…”

“…Joint space width data were utilized by Cheung et al [49] to feed into the XGBoost model for knee OA severity classification.…”

Comparison of Radiologic Results after Lateral Meniscal Allograft Transplantation with or without Capsulodesis Using an All-Soft Suture Anchor

“…We did not apply universal classification such as Kellgren-Lawrence grade to determine joint space narrowing of the lateral edge [ 39 , 40 ]. Lee et al [ 24 ] revealed that joint cartilage is relatively well maintained when only the lateral edge, not the joint obliteration, is narrow in a considerable number of patients.…”

Quantifying Joint Congruence With an Elastic Foundation

“…Similarly, the was no relationship evident in the ten joints between R e and mJSW. However, studies have found fixed location and multiple location measures of joint space width to be superior to mJSW [45,46]. Thus, it is possible this lack of correlation was due in part to the limitations of mJSW as a parameter.…”