B C S de Vries1, J H Hegeman2,3, W Nijmeijer2, J Geerdink4, C Seifert3, C G M Groothuis-Oudshoorn5. 1. Department of Trauma Surgery, Ziekenhuisgroep Twente, Zilvermeeuw 1, 7609 PP, Almelo, The Netherlands. bcsdevries@gmail.com. 2. Department of Trauma Surgery, Ziekenhuisgroep Twente, Zilvermeeuw 1, 7609 PP, Almelo, The Netherlands. 3. Faculty of Electrical Engineering, Mathematics and Computer Science, University of Twente, Enschede, The Netherlands. 4. Department of Information & Organization, Ziekenhuisgroep Twente, Almelo, The Netherlands. 5. Health Technology and Services Research, Techmed Centre, University of Twente, Enschede, The Netherlands.
Abstract
Four machine learning models were developed and compared to predict the risk of a future major osteoporotic fracture (MOF), defined as hip, wrist, spine and humerus fractures, in patients with a prior fracture. We developed a user-friendly tool for risk calculation of subsequent MOF in osteopenia patients, using the best performing model. INTRODUCTION: Major osteoporotic fractures (MOFs), defined as hip, wrist, spine and humerus fractures, can have serious consequences regarding morbidity and mortality. Machine learning provides new opportunities for fracture prediction and may aid in targeting preventive interventions to patients at risk of MOF. The primary objective is to develop and compare several models, capable of predicting the risk of MOF as a function of time in patients seen at the fracture and osteoporosis outpatient clinic (FO-clinic) after sustaining a fracture. METHODS: Patients aged > 50 years visiting an FO-clinic were included in this retrospective study. We compared discriminative ability (concordance index) for predicting the risk on MOF with a Cox regression, random survival forests (RSF) and an artificial neural network (ANN)-DeepSurv model. Missing data was imputed using multiple imputations by chained equations (MICE) or RSF's imputation function. Analyses were performed for the total cohort and a subset of osteopenia patients without vertebral fracture. RESULTS: A total of 7578 patients were included, 805 (11%) patients sustained a subsequent MOF. The highest concordance-index in the total dataset was 0.697 (0.664-0.730) for Cox regression; no significant difference was determined between the models. In the osteopenia subset, Cox regression outperformed RSF (p = 0.043 and p = 0.023) and ANN-DeepSurv (p = 0.043) with a c-index of 0.625 (0.562-0.689). Cox regression was used to develop a MOF risk calculator on this subset. CONCLUSION: We show that predicting the risk of MOF in patients who already sustained a fracture can be done with adequate discriminative performance. We developed a user-friendly tool for risk calculation of subsequent MOF in patients with osteopenia.
Four machine learning models were developed and compared to predict the risk of a future major osteoporotic fracture (MOF), defined as hip, wrist, spine and humerus fractures, in patients with a prior fracture. We developed a user-friendly tool for risk calculation of subsequent MOF in osteopeniapatients, using the best performing model. INTRODUCTION: Major osteoporotic fractures (MOFs), defined as hip, wrist, spine and humerus fractures, can have serious consequences regarding morbidity and mortality. Machine learning provides new opportunities for fracture prediction and may aid in targeting preventive interventions to patients at risk of MOF. The primary objective is to develop and compare several models, capable of predicting the risk of MOF as a function of time in patients seen at the fracture and osteoporosisoutpatient clinic (FO-clinic) after sustaining a fracture. METHODS:Patients aged > 50 years visiting an FO-clinic were included in this retrospective study. We compared discriminative ability (concordance index) for predicting the risk on MOF with a Cox regression, random survival forests (RSF) and an artificial neural network (ANN)-DeepSurv model. Missing data was imputed using multiple imputations by chained equations (MICE) or RSF's imputation function. Analyses were performed for the total cohort and a subset of osteopeniapatients without vertebral fracture. RESULTS: A total of 7578 patients were included, 805 (11%) patients sustained a subsequent MOF. The highest concordance-index in the total dataset was 0.697 (0.664-0.730) for Cox regression; no significant difference was determined between the models. In the osteopenia subset, Cox regression outperformed RSF (p = 0.043 and p = 0.023) and ANN-DeepSurv (p = 0.043) with a c-index of 0.625 (0.562-0.689). Cox regression was used to develop a MOF risk calculator on this subset. CONCLUSION: We show that predicting the risk of MOF in patients who already sustained a fracture can be done with adequate discriminative performance. We developed a user-friendly tool for risk calculation of subsequent MOF in patients with osteopenia.
Authors: M Jürisson; H Pisarev; J Kanis; F Borgström; A Svedbom; R Kallikorm; M Lember; A Uusküla Journal: Osteoporos Int Date: 2016-02-23 Impact factor: 4.507
Authors: J H Hegeman; G Willemsen; J van Nieuwpoort; H G Kreeftenberg; E van der Veer; J P J Slaets; H J ten Duis Journal: Ned Tijdschr Geneeskd Date: 2004-10-30
Authors: Karine Briot; Simon Paternotte; Sami Kolta; Richard Eastell; Dieter Felsenberg; David M Reid; Claus-C Glüer; Christian Roux Journal: PLoS One Date: 2013-12-30 Impact factor: 3.240
Authors: Jared L Katzman; Uri Shaham; Alexander Cloninger; Jonathan Bates; Tingting Jiang; Yuval Kluger Journal: BMC Med Res Methodol Date: 2018-02-26 Impact factor: 4.615
Authors: Babak Saravi; Frank Hassel; Sara Ülkümen; Alisia Zink; Veronika Shavlokhova; Sebastien Couillard-Despres; Martin Boeker; Peter Obid; Gernot Michael Lang Journal: J Pers Med Date: 2022-03-22
Authors: Bomee Kim; Yun Ji Jang; Hae Ram Cho; So Yeon Kim; Ji Eun Jeong; Mi Kyoung Shim; Myeong Gyu Kim Journal: Clin Transl Sci Date: 2021-11-17 Impact factor: 4.689