Publicação
Enhancing osteoporosis risk prediction using machine learning: a holistic approach integrating biomarkers and clinical data
| datacite.subject.sdg | 03:Saúde de Qualidade | |
| datacite.subject.sdg | 09:Indústria, Inovação e Infraestruturas | |
| datacite.subject.sdg | 10:Reduzir as Desigualdades | |
| dc.contributor.author | Pires de Carvalho, Filipe Ricardo | |
| dc.contributor.author | Gavaia, Paulo | |
| dc.date.accessioned | 2026-04-21T09:32:33Z | |
| dc.date.available | 2026-04-21T09:32:33Z | |
| dc.date.issued | 2025-06 | |
| dc.description.abstract | Osteoporosis (OP) affects approximately 18 % of the global population, with osteoporosis-associated fractures impacting up to 37 million people annually. While dual-energy X-ray absorptiometry (DXA) remains the gold standard for diagnosis, its limitations, including restricted availability and radiation exposure, highlight the need for alternative screening methods. We developed a machine learning model to predict OP risk using routinely collected clinical data, deliberately excluding DXA measurements to ensure broad accessibility. Using data from NHANES cycles 2007–2014, we analyzed 7924 participants aged 50 years and older, identifying 1636 OP cases (20.6 %) and 6288 normal cases (79.4 %) through comprehensive criteria incorporating both WHO densitometric standards (T-scores ≤ − 2.5) and anthropometric risk factors. We implemented a stacking ensemble model combining four specialized classifiers (Gradient Boosting, Random Forest, XGBoost, and LightGBM) with a logistic regression meta-classifier. The model achieved 93 % accuracy, an AUC of 0.94, and demonstrated robust performance through cross-validation (mean score: 0.929 ± 0.030). feature importance analysis revealed age (6.04 %), arm muscle circumference (5.61 %), and body weight (5.30 %) as the most influential predictors, followed by gender (3.28 %), BMI (2.71 %), and calcium intake (2.42 %). Additional significant predictors included folate (2.28 %), height (2.23 %), hand grip strength (2.21 %), and alkaline phosphatase (2.16 %). These biologically plausible relationships align with established clinical knowledge of OP risk factors. The model’s strong performance metrics and reliance on readily available clinical data suggest its potential as a practical screening tool, particularly in settings with limited DXA access. All code and implementation details are openly available on GitHub, facilitating integration into existing healthcare systems. This approach offers a promising pathway for enhancing early OP detection and risk assessment across diverse healthcare settings. | eng |
| dc.identifier.doi | 10.1016/j.compbiomed.2025.110289 | |
| dc.identifier.issn | 0010-4825 | |
| dc.identifier.uri | http://hdl.handle.net/10400.1/28723 | |
| dc.language.iso | eng | |
| dc.peerreviewed | yes | |
| dc.publisher | Elsevier | |
| dc.relation | Algarve Centre for Marine Sciences | |
| dc.relation | Algarve Centre for Marine Sciences | |
| dc.relation | Centre for Marine and Environmental Research | |
| dc.relation.ispartof | Computers in Biology and Medicine | |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | |
| dc.subject | Machine learning | |
| dc.subject | Osteoporosis risk prediction | |
| dc.subject | Biomarkers | |
| dc.subject | NHANES | |
| dc.subject | Stacking ensemble | |
| dc.subject | Bone mineral density | |
| dc.subject | Preventive medicine | |
| dc.title | Enhancing osteoporosis risk prediction using machine learning: a holistic approach integrating biomarkers and clinical data | eng |
| dc.type | journal article | |
| dspace.entity.type | Publication | |
| oaire.awardNumber | UIDB/04326/2020 | |
| oaire.awardNumber | UIDP/04326/2020 | |
| oaire.awardNumber | LA/P/0101/2020 | |
| oaire.awardTitle | Algarve Centre for Marine Sciences | |
| oaire.awardTitle | Algarve Centre for Marine Sciences | |
| oaire.awardTitle | Centre for Marine and Environmental Research | |
| oaire.awardURI | info:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F04326%2F2020/PT | |
| oaire.awardURI | info:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDP%2F04326%2F2020/PT | |
| oaire.awardURI | info:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/LA%2FP%2F0101%2F2020/PT | |
| oaire.citation.startPage | 110289 | |
| oaire.citation.title | Computers in Biology and Medicine | |
| oaire.citation.volume | 192 | |
| oaire.fundingStream | 6817 - DCRRNI ID | |
| oaire.fundingStream | 6817 - DCRRNI ID | |
| oaire.fundingStream | 6817 - DCRRNI ID | |
| oaire.version | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |
| person.familyName | Pires de Carvalho | |
| person.familyName | Gavaia | |
| person.givenName | Filipe Ricardo | |
| person.givenName | Paulo | |
| person.identifier.ciencia-id | 181A-A440-7D6E | |
| person.identifier.ciencia-id | B619-FC16-D007 | |
| person.identifier.orcid | 0000-0002-1468-0305 | |
| person.identifier.orcid | 0000-0002-9582-1957 | |
| person.identifier.rid | A-6470-2011 | |
| person.identifier.scopus-author-id | 55115644400 | |
| person.identifier.scopus-author-id | 6507104377 | |
| project.funder.identifier | http://doi.org/10.13039/501100001871 | |
| project.funder.identifier | http://doi.org/10.13039/501100001871 | |
| project.funder.identifier | http://doi.org/10.13039/501100001871 | |
| project.funder.name | Fundação para a Ciência e a Tecnologia | |
| project.funder.name | Fundação para a Ciência e a Tecnologia | |
| project.funder.name | Fundação para a Ciência e a Tecnologia | |
| relation.isAuthorOfPublication | 55c2b5fb-f39a-4675-834a-f44c5fe417aa | |
| relation.isAuthorOfPublication | 9dca2139-21a4-4d59-aaf7-531f1033a58e | |
| relation.isAuthorOfPublication.latestForDiscovery | 55c2b5fb-f39a-4675-834a-f44c5fe417aa | |
| relation.isProjectOfPublication | fafa76a6-2cd2-4a6d-a3c9-772f34d3b91f | |
| relation.isProjectOfPublication | 15f91d45-e070-47d8-b6b8-efd4de31d9a8 | |
| relation.isProjectOfPublication | 794d4c77-c731-471e-bc96-5a41dcd3d872 | |
| relation.isProjectOfPublication.latestForDiscovery | fafa76a6-2cd2-4a6d-a3c9-772f34d3b91f |