{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T03:14:58Z","timestamp":1789614898562,"version":"build-2803163510"},"reference-count":23,"publisher":"Georg Thieme Verlag KG","issue":"01\/02","license":[{"start":{"date-parts":[[2024,4,11]],"date-time":"2024-04-11T00:00:00Z","timestamp":1712793600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Methods Inf Med"],"published-print":{"date-parts":[[2024,5]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>\n          Objective\u2003In this study, we propose a novel framework that utilizes deep learning and attention mechanisms to predict the radiographic progression of patellofemoral osteoarthritis (PFOA) over a period of 7 years.<\/jats:p><jats:p>\n          Material and Methods\u2003This study included subjects (1,832 subjects, 3,276 knees) from the baseline of the Multicenter Osteoarthritis Study (MOST). Patellofemoral joint regions of interest were identified using an automated landmark detection tool (BoneFinder) on lateral knee X-rays. An end-to-end deep learning method was developed for predicting PFOA progression based on imaging data in a five-fold cross-validation setting. To evaluate the performance of the models, a set of baselines based on known risk factors were developed and analyzed using gradient boosting machine (GBM). Risk factors included age, sex, body mass index, and Western Ontario and McMaster Universities Arthritis Index score, and the radiographic osteoarthritis stage of the tibiofemoral joint (Kellgren and Lawrence [KL] score). Finally, to increase predictive power, we trained an ensemble model using both imaging and clinical data.<\/jats:p><jats:p>\n          Results\u2003Among the individual models, the performance of our deep convolutional neural network attention model achieved the best performance with an area under the receiver operating characteristic curve (AUC) of 0.856 and average precision (AP) of 0.431, slightly outperforming the deep learning approach without attention (AUC\u2009=\u20090.832, AP\u2009=\u20090.4) and the best performing reference GBM model (AUC\u2009=\u20090.767, AP\u2009=\u20090.334). The inclusion of imaging data and clinical variables in an ensemble model allowed statistically more powerful prediction of PFOA progression (AUC\u2009=\u20090.865, AP\u2009=\u20090.447), although the clinical significance of this minor performance gain remains unknown. The spatial attention module improved the predictive performance of the backbone model, and the visual interpretation of attention maps focused on the joint space and the regions where osteophytes typically occur.<\/jats:p><jats:p>\n          Conclusion\u2003This 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.<\/jats:p>","DOI":"10.1055\/a-2305-2115","type":"journal-article","created":{"date-parts":[[2024,4,11]],"date-time":"2024-04-11T23:20:12Z","timestamp":1712877612000},"page":"001-010","source":"Crossref","is-referenced-by-count":9,"title":["Deep Learning for Predicting Progression of Patellofemoral Osteoarthritis Based on Lateral Knee Radiographs, Demographic Data, and Symptomatic Assessments"],"prefix":"10.1055","volume":"63","author":[{"given":"Neslihan","family":"Bayramoglu","sequence":"additional","affiliation":[{"name":"Research Unit of Health Sciences and Technology, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Martin","family":"Englund","sequence":"additional","affiliation":[{"name":"Orthopaedics, Department of Clinical Sciences Lund Faculty of Medicine, Lund University, Lund, Sweden"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ida K.","family":"Haugen","sequence":"additional","affiliation":[{"name":"Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Muneaki","family":"Ishijima","sequence":"additional","affiliation":[{"name":"Department of Orthopaedics, Faculty of Medicine, Juntendo University, Tokyo, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Simo","family":"Saarakkala","sequence":"additional","affiliation":[{"name":"Research Unit of Health Sciences and Technology, University of Oulu, Oulu, Finland"},{"name":"Department of Diagnostic Radiology, Oulu University Hospital, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"194","published-online":{"date-parts":[[2024,5,14]]},"reference":[{"issue":"05","key":"ref1","doi-asserted-by":"crossref","first-page":"647","DOI":"10.1016\/j.joca.2016.12.006","article-title":"Incidence, prevalence, natural course and prognosis of patellofemoral osteoarthritis: the Cohort Hip and Cohort Knee study","volume":"25","author":"N E Lankhorst","year":"2017","journal-title":"Osteoarthritis Cartilage"},{"issue":"11","key":"ref2","doi-asserted-by":"crossref","first-page":"1704","DOI":"10.1093\/rheumatology\/ken339","article-title":"How do pain and function vary with compartmental distribution and severity of radiographic knee osteoarthritis?","volume":"47","author":"R Duncan","year":"2008","journal-title":"Rheumatology (Oxford)"},{"issue":"07","key":"ref3","doi-asserted-by":"crossref","first-page":"765","DOI":"10.1016\/j.joca.2011.05.005","article-title":"The patellofemoral joint: the forgotten joint in knee osteoarthritis","volume":"19","author":"K M Crossley","year":"2011","journal-title":"Osteoarthritis Cartilage"},{"issue":"11","key":"ref4","doi-asserted-by":"crossref","first-page":"1944","DOI":"10.1136\/ard.2011.151050","article-title":"Incidence, progression and sequence of development of radiographic knee osteoarthritis in a symptomatic population","volume":"70","author":"R Duncan","year":"2011","journal-title":"Ann Rheum Dis"},{"issue":"09","key":"ref5","doi-asserted-by":"crossref","first-page":"1151","DOI":"10.1016\/j.joca.2009.03.016","article-title":"Does isolated patellofemoral osteoarthritis matter?","volume":"17","author":"R Duncan","year":"2009","journal-title":"Osteoarthritis Cartilage"},{"issue":"04","key":"ref6","doi-asserted-by":"crossref","first-page":"193","DOI":"10.5792\/ksrr.2012.24.4.193","article-title":"Patellofemoral osteoarthritis","volume":"24","author":"Y M Kim","year":"2012","journal-title":"Knee Surg Relat Res"},{"key":"ref7","doi-asserted-by":"crossref","first-page":"666","DOI":"10.1016\/j.semarthrit.2017.09.009","article-title":"International patellofemoral osteoarthritis consortium: consensus statement on the diagnosis, burden, outcome measures, prognosis, risk factors and treatment","volume":"47","author":"M van Middelkoop","year":"2018","journal-title":"Semin Arthritis Rheum"},{"issue":"10","key":"ref8","doi-asserted-by":"crossref","first-page":"1697","DOI":"10.1016\/j.joca.2016.05.011","article-title":"The prevalence of patellofemoral osteoarthritis: a systematic review and meta-analysis","volume":"24","author":"S Kobayashi","year":"2016","journal-title":"Osteoarthritis Cartilage"},{"key":"ref9","volume-title":"Patellofemoral osteoarthritis: characterizing knee alignment and morphology [PhD thesis]","author":"E M Macri","year":"2017"},{"issue":"02","key":"ref10","doi-asserted-by":"crossref","first-page":"R63","DOI":"10.1186\/ar3779","article-title":"Clinical features of symptomatic patellofemoral joint osteoarthritis","volume":"14","author":"G Peat","year":"2012","journal-title":"Arthritis Res Ther"},{"issue":"03","key":"ref11","doi-asserted-by":"crossref","first-page":"473","DOI":"10.1016\/j.joca.2015.09.021","article-title":"Radiographic progression of knee osteoarthritis is associated with MRI abnormalities in both the patellofemoral and tibiofemoral joint","volume":"24","author":"B JE de Lange-Brokaar","year":"2016","journal-title":"Osteoarthritis Cartilage"},{"issue":"10","key":"ref12","doi-asserted-by":"crossref","first-page":"1432","DOI":"10.1016\/j.joca.2021.06.011","article-title":"Automated detection of patellofemoral osteoarthritis from knee lateral view radiographs using deep learning: data from the Multicenter Osteoarthritis Study (MOST)","volume":"29","author":"N Bayramoglu","year":"2021","journal-title":"Osteoarthritis Cartilage"},{"key":"ref13","doi-asserted-by":"crossref","first-page":"104627","DOI":"10.1016\/j.ijmedinf.2021.104627","article-title":"Machine learning based texture analysis of patella from X-rays for detecting patellofemoral osteoarthritis","volume":"157","author":"N Bayramoglu","year":"2022","journal-title":"Int J Med Inform"},{"issue":"08","key":"ref14","doi-asserted-by":"crossref","first-page":"1462","DOI":"10.1109\/TMI.2013.2258030","article-title":"Fully automatic segmentation of the proximal femur using random forest regression voting","volume":"32","author":"C Lindner","year":"2013","journal-title":"IEEE Trans Med Imaging"},{"key":"ref15","doi-asserted-by":"crossref","first-page":"1189","DOI":"10.1214\/aos\/1013203451","article-title":"Greedy function approximation: a gradient boosting machine","volume":"29","author":"J H Friedman","year":"2001","journal-title":"Ann Statist"},{"issue":"06","key":"ref16","doi-asserted-by":"crossref","first-page":"748","DOI":"10.1016\/j.joca.2008.09.013","article-title":"The association of meniscal damage with joint effusion in persons without radiographic osteoarthritis: the Framingham and MOST osteoarthritis studies","volume":"17","author":"F W Roemer","year":"2009","journal-title":"Osteoarthritis Cartilage"},{"key":"ref22","first-page":"30","article-title":"Attention is all you need","author":"A Vaswani","year":"2017","journal-title":"Adv Neural Inf Process Syst"},{"key":"ref23","first-page":"3146","volume-title":"LightGBM: A highly efficient gradient boosting decision tree","author":"G Ke","year":"2017"},{"key":"ref24","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1175\/1520-0493(1950)078<0001:VOFEIT>2.0.CO;2","article-title":"Verification of forecasts expressed in terms of probability","volume":"78","author":"G W Brier","year":"1950","journal-title":"Mon Weather Rev"},{"issue":"03","key":"ref25","doi-asserted-by":"crossref","first-page":"837","DOI":"10.2307\/2531595","article-title":"Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach","volume":"44","author":"E R DeLong","year":"1988","journal-title":"Biometrics"},{"key":"ref26","first-page":"4765","volume-title":"A unified approach to interpreting model predictions","author":"S M Lundberg","year":"2017"},{"issue":"01","key":"ref27","doi-asserted-by":"crossref","first-page":"20038","DOI":"10.1038\/s41598-019-56527-3","article-title":"Multimodal machine learning-based knee osteoarthritis progression prediction from plain radiographs and clinical data","volume":"9","author":"A Tiulpin","year":"2019","journal-title":"Sci Rep"},{"key":"ref28","doi-asserted-by":"crossref","first-page":"48","DOI":"10.1016\/j.neucom.2021.03.091","article-title":"A review on the attention mechanism of deep learning","volume":"452","author":"Z Niu","year":"2021","journal-title":"Neurocomputing"}],"container-title":["Methods of Information in Medicine"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/www.thieme-connect.de\/products\/ejournals\/pdf\/10.1055\/a-2305-2115.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,11,5]],"date-time":"2024-11-05T11:51:15Z","timestamp":1730807475000},"score":1,"resource":{"primary":{"URL":"http:\/\/www.thieme-connect.de\/DOI\/DOI?10.1055\/a-2305-2115"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,4,11]]},"references-count":23,"journal-issue":{"issue":"01\/02","published-online":{"date-parts":[[2024,10,22]]},"published-print":{"date-parts":[[2024,5]]}},"URL":"https:\/\/doi.org\/10.1055\/a-2305-2115","archive":["Portico","CLOCKSS"],"relation":{},"ISSN":["0026-1270","2511-705X"],"issn-type":[{"value":"0026-1270","type":"print"},{"value":"2511-705X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2024,4,11]]},"published":{"date-parts":[[2024,4,11]]}}}