Dataset Information

Osteoporosis Pre-Screening Using Ensemble Machine Learning in Postmenopausal Korean Women.

ABSTRACT: As osteoporosis is a degenerative disease related to postmenopausal aging, early diagnosis is vital. This study used data from the Korea National Health and Nutrition Examination Surveys to predict a patient's risk of osteoporosis using machine learning algorithms. Data from 1431 postmenopausal women aged 40-69 years were used, including 20 features affecting osteoporosis, chosen by feature importance and recursive feature elimination. Random Forest (RF), AdaBoost, and Gradient Boosting (GBM) machine learning algorithms were each used to train three models: A, checkup features; B, survey features; and C, both checkup and survey features, respectively. Of the three models, Model C generated the best outcomes with an accuracy of 0.832 for RF, 0.849 for AdaBoost, and 0.829 for GBM. Its area under the receiver operating characteristic curve (AUROC) was 0.919 for RF, 0.921 for AdaBoost, and 0.908 for GBM. By utilizing multiple feature selection methods, the ensemble models of this study achieved excellent results with an AUROC score of 0.921 with AdaBoost, which is 0.1-0.2 higher than those of the best performing models from recent studies. Our model can be further improved as a practical medical tool for the early diagnosis of osteoporosis after menopause.

SUBMITTER: Kwon Y

PROVIDER: S-EPMC9222287 | biostudies-literature | 2022 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Osteoporosis Pre-Screening Using Ensemble Machine Learning in Postmenopausal Korean Women.

Kwon Youngihn Y Lee Juyeon J Park Joo Hee JH Kim Yoo Mee YM Kim Se Hwa SH Won Young Jun YJ Kim Hyung-Yong HY

Healthcare (Basel, Switzerland) 20220614 6

As osteoporosis is a degenerative disease related to postmenopausal aging, early diagnosis is vital. This study used data from the Korea National Health and Nutrition Examination Surveys to predict a patient's risk of osteoporosis using machine learning algorithms. Data from 1431 postmenopausal women aged 40-69 years were used, including 20 features affecting osteoporosis, chosen by feature importance and recursive feature elimination. Random Forest (RF), AdaBoost, and Gradient Boosting (GBM) ma ...[more]

PMID: 35742158

Similar Datasets

Project description:Osteoporosis is a disease characterized by impairment of bone microarchitecture that causes high socioeconomic impacts in the world because of fractures and hospitalizations. Although dual-energy X-ray absorptiometry (DXA) is the gold standard for diagnosing the disease, access to DXA in developing countries is still limited due to its high cost, being present only in specialized hospitals. In this paper, we analyze the performance of Osseus, a low-cost portable device based on electromagnetic waves that measures the attenuation of the signal that crosses the medial phalanx of a patient's middle finger and was developed for osteoporosis screening. The analysis is carried out by predicting changes in bone mineral density using Osseus measurements and additional common risk factors used as input features to a set of supervised classification models, while the results from DXA are taken as target (real) values during the training of the machine learning algorithms. The dataset consisted of 505 patients who underwent osteoporosis screening with both devices (DXA and Osseus), of whom 21.8% were healthy and 78.2% had low bone mineral density or osteoporosis. A cross-validation with k-fold = 5 was considered in model training, while 20% of the whole dataset was used for testing. The obtained performance of the best model (Random Forest) presented a sensitivity of 0.853, a specificity of 0.879, and an F1 of 0.859. Since the Random Forest (RF) algorithm allows some interpretability of its results (through the impurity check), we were able to identify the most important variables in the classification of osteoporosis. The results showed that the most important variables were age, body mass index, and the signal attenuation provided by Osseus. The RF model, when used together with Osseus measurements, is effective in screening patients and facilitates the early diagnosis of osteoporosis. The main advantages of such early screening are the reduction of costs associated with exams, surgeries, treatments, and hospitalizations, as well as improved quality of life for patients.

Project description:ImportanceFor younger postmenopausal women, clinical guidelines recommend using osteoporosis risk prediction tools to identify candidates with low bone mineral density (BMD). However, the performance of these tools is not well quantified.ObjectiveTo examine the performance of Osteoporosis Risk Assessment Instrument (ORAI) and Osteoporosis Index of Risk (OSIRIS), compared with Osteoporosis Self-Assessment Tool (OST), in identifying the presence of osteoporotic BMD in younger postmenopausal women.Design, setting, and participantsThis cross-sectional study used data from the Women's Health Initiative Bone Density Substudy, which was conducted at 3 clinical centers in Tucson and Phoenix, Arizona; Pittsburgh, Pennsylvania; and Birmingham, Alabama. Participants were healthy postmenopausal women aged 50 to 64 years with BMD measurements evaluated using the 3 risk prediction tools: OSIRIS, ORAI, and OST. Risk factors and other participant characteristics were compared across osteoporosis status. Data were collected from October 1993 to December 1998 and analyzed between September 23, 2023, and April 10, 2024.ExposuresThe primary exposures were OSIRIS, ORAI, and OST risk scores.Main outcomes and measuresPrimary outcome was osteoporosis defined by BMD T score of -2.5 or lower at 1 or more of 3 anatomical locations: femoral neck, total hip, and/or lumbar spine. The tools were evaluated via area under the receiver operating characteristic curve (AUC) at published score cutoffs and at alternate cutoffs.ResultsAmong 6067 included participants (mean [SD] age at baseline, 57.7 [4.1] years), the prevalence of osteoporosis was 14.1% (n = 857) at any 1 of 3 anatomical sites. AUC for identifying osteoporosis at any site was 0.633 (95% CI, 0.633-0.634) for OSIRIS, 0.663 (95% CI, 0.663-0.664) for ORAI, and 0.654 (95% CI, 0.654-0.655) for OST.Conclusions and relevanceIn this cross-sectional study, 3 guideline-recommended osteoporosis risk assessment tools had fair to moderate discrimination in identifying osteoporosis defined by lowest BMD at any 1 of 3 skeletal sites. Screening is essential to reducing individual and societal burden of osteoporosis and related fractures, and this study showed a gap in identifying younger postmenopausal women using common clinical risk factors.

Project description:The optimal approach to osteoporosis screening and treatment in postmenopausal women is unclear. We compared (i) the United States Preventive Services Task Force (USPSTF) and Osteoporosis Canada osteoporosis screening strategies; and (ii) the National Osteoporosis Foundation (NOF) and Canadian treatment strategies. We used data from the prospective Women's Health Initiative Observational Study and Clinical Trials of women aged 50 to 79 years at baseline (n = 117,707 followed for self-reported fractures; n = 8134 in bone mineral density [BMD] subset). We determined the yield of the screening and treatment strategies in identifying women who experienced major osteoporotic fractures (MOFs) during a 10-year follow-up. Among women aged 50 to 64 years, 23.1% of women were identified for BMD testing under the USPSTF strategy and 52.3% under the Canadian strategy. For women ≥65 years, 100% were identified for testing under the USPSTF and Canadian strategies, 35% to 74% were identified for treatment under NOF, and 16% to 37% were identified for treatment under CAROC (range among 5-year age subgroups). Among women who experienced MOF during follow-up, the USPSTF strategy identified 6.7% of women 50 to 54 years-old and 49.5% of women 60 to 64 years-old for BMD testing (versus 54.4% and 60.6% for the Canadian strategy, respectively). However, the specificity of the USPSTF strategy was higher than that of the Canadian strategy among women 50 to 64 years-old. Among women who experienced MOF during follow-up, sensitivity for identifying women as treatment candidates was lowest for both strategies in women aged 50 to 64 (NOF 10% to 38%; CAROC 1% to 15%) and maximal in 75-year-old to 79-year-old women (NOF 82.8%; 51.6% CAROC); specificity declined with advancing age and was lower with the NOF compared to the CAROC strategy. Among women aged 50 to 64 years, the screening and treatment strategies examined had low sensitivity for identifying those who subsequently experience MOF; sensitivity was higher among women ≥65 years than among younger women. New screening and treatment algorithms are needed. © 2018 American Society for Bone and Mineral Research.

Dataset Information

Osteoporosis Pre-Screening Using Ensemble Machine Learning in Postmenopausal Korean Women.

Publications

Osteoporosis Pre-Screening Using Ensemble Machine Learning in Postmenopausal Korean Women.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets