2021/06/03 by Neslihan Bayramoğlu, Bayramoglu, Neslihan, Miika T. Nieminen +3
Engineering · Medicine · #Computer Vision and Pattern Recognition (cs.CV) #Diabetic Foot Ulcer Assessment and Management #FOS: Computer and information sciences #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Lower Extremity Biomechanics and Pathologies #Machine Learning (cs.LG) #Osteoarthritis Treatment and Mechanisms #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2106.01700
openalex publication_date 2021/06/03 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Objective is to assess the ability of texture features for detecting\nradiographic patellofemoral osteoarthritis (PFOA) from knee lateral view\nradiographs. We used lateral view knee radiographs from MOST public use\ndatasets (n = 5507 knees). Patellar region-of-interest (ROI) was automatically\ndetected using landmark detection tool (BoneFinder). Hand-crafted features,\nbased on LocalBinary Patterns (LBP), were then extracted to describe the\npatellar texture. First, a machine learning model (Gradient Boosting Machine)\nwas trained to detect radiographic PFOA from the LBP features. Furthermore, we\nused end-to-end trained deep convolutional neural networks (CNNs) directly on\nthe texture patches for detecting the PFOA. The proposed classification models\nwere eventually compared with more conventional reference models that use\nclinical assessments and participant characteristics such as age, sex, body\nmass index(BMI), the total WOMAC score, and tibiofemoral Kellgren-Lawrence (KL)\ngrade. Atlas-guided visual assessment of PFOA status by expert readers provided\nin the MOST public use datasets was used as a classification outcome for the\nmodels. Performance of prediction models was assessed using the area under the\nreceiver operating characteristic curve (ROC AUC), the area under the\nprecision-recall (PR) curve-average precision (AP)-, and Brier score in the\nstratified 5-fold cross validation setting.Of the 5507 knees, 953 (17.3%) had\nPFOA. AUC and AP for the strongest reference model including age, sex, BMI,\nWOMAC score, and tibiofemoral KL grade to predict PFOA were 0.817 and 0.487,\nrespectively. Textural ROI classification using CNN significantly improved the\nprediction performance (ROC AUC= 0.889, AP= 0.714). We present the first study\nthat analyses patellar bone texture for diagnosing PFOA. Our results\ndemonstrates the potential of using texture features of patella to predict\nPFOA.\n