Osteoporosis screening using machine learning and electromagnetic waves - Research and performance evaluation

Aug 5, 2026

Upload by Patrick

Osteoporosis causes impaired bone microarchitecture and leads to fractures and hospitalizations with substantial socioeconomic burden worldwide. Although dual-energy X-ray absorptiometry (DXA) is the diagnostic gold standard, limited access in developing regions is driven by cost and infrastructure constraints. The work evaluates Osseus, a low-cost portable electromagnetic-wave device, predicting changes in bone mineral density using supervised classification models with Osseus attenuation plus common risk factors, trained against DXA targets. Using 5-fold cross-validation and a 20% test split, the best Random Forest model achieves sensitivity 0.853, specificity 0.879, and F1 0.859. Feature analysis identifies age, body mass index, and signal attenuation as key variables, supporting effective early screening and reduced diagnostic costs.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Background and need for low-cost osteoporosis diagnosis

Gold-standard DXA limitations

Osteoporosis risk factors and clinical relevance

Proposed solution: Osseus device and modeling pipeline

Electromagnetic-wave attenuation measurement

Feature design with risk factors and device signals

Supervised learning with DXA targets

Dataset and evaluation strategy

Patient cohort and class distribution

Cross-validation and testing protocol

Results and interpretation

Best model performance metrics

Variable importance and clinical implications

Advantages of early screening

[www. nature.com/scientificreports](www. nature.com/scientificreports) OPEN Osteoporosis screening using machine learning and electromagnetic waves GabrielaA. Albuquerque1,2*, Dionísio D. A. Carvalho1,2, Agnaldo S. Cruz1,2, João P. Q. Santos2, Guilherme M. Machado3, Ignácio S. Gendriz1, Felipe R. S. Fernandes1, Ingridy M. P. Barbalho1, Marquiony M. Santos1, CésarA. D. Teixeira4, Jorge M. O. Henriques4, Paulo Gil5, Adrião D. D. Neto6, Antonio L. P. S. Campos6, Josivan G. Lima7, Jailton C. Paiva2, Antonio H. F. Morais2, Thaisa Santos Lima1,8 & Ricardo A. M. Valentim1 Osteoporosis is a disease characterized by impairment of bone microarchitecture that causes high socioeconomic impacts in the world because of fractures and hospitalizations. Although dualenergy 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 ofOsseus, 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. Across-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. Osteoporosis is characterized by impaired bone strength1 and affects approximately 6.3% of men over the age of 50 and 21.2% of women over the same age range globally, i.e., approximately 500 million men and women worldwide2. The disease causes more than 8.9 million fractures annually, resulting in one osteoporosis fracture every 3 s3. In the US, the annual direct medical cost of osteoporosis in 2005 was 17 billion USD and is projected to rise to 25 billion USD by 20254. In the 27 countries of the European Union, the estimated cost was 34.5 billion dollars in 2010. In four latin american countries (Brazil, Mexico, Colombia, and Argentina), the burden of the disease in 2018 was estimated at 1.17 billion dollars5. In Brazil, a 63% increase in the annual number of fractures 1Laboratory of Technological Innovation in Health (LAIS), Natal, RN, Brazil. 2Advanced Nucleus of Technological Innovation (NAVI), Federal Institute of Rio Grande do Norte (IFRN), Natal, RN, Brazil. 3LyRIDS, ECE-Engineering School, Paris, France. 4Department of Informatics Engineering, Univ. Coimbra, Centre for Informatics and Systems of the University of Coimbra (CISUC), Coimbra, Portugal. 5Department of Electrical and Computer Engineering, School of Science and Technology, New University of Lisbon, Lisbon, Portugal. 6Post-Gradu

FAQs

Why is osteoporosis screening needed and what impact does the disease have?

Osteoporosis impairs bone microarchitecture and drives high socioeconomic burden through fractures and hospitalizations. It is linked to millions of fractures annually and increasing direct medical costs.

How does the proposed Osseus approach differ from DXA?

DXA is the gold standard but requires expensive infrastructure and specialized operation. Osseus is a low-cost portable electromagnetic-wave device that measures signal attenuation across the patient’s finger and supports prediction models trained against DXA.

Which machine learning model performed best and what were the key results?

The Random Forest model achieved sensitivity 0.853, specificity 0.879, and F1 0.859 under 5-fold cross-validation with a 20% test split. Variable interpretation highlighted age, body mass index, and Osseus signal attenuation as most important.

Aug 5, 2026

Upload by Patrick