Related Experiment Videos
Artificial neural network analysis: a novel application for predicting site-specific bone mineral density.
E I Mohamed1, C Maiolo, R Linder
1Department of Neurosciences Building F-Sud, 1st Floor, Room 116, University of Tor Vergata, Via Montpellier 1, I-00133, Rome, Italy. eimohamed@med.uniroma2.it
Acta Diabetologica
|November 18, 2003
Summary
Artificial neural networks (ANN) can accurately estimate bone mineral density (BMD) using simple anthropometric measurements. This approach offers a promising, cost-effective alternative to dual X-ray absorptiometry (DXA) for bone health screening.
Area of Science:
- Biomedical Engineering
- Medical Imaging
- Data Science
Background:
- Dual X-ray absorptiometry (DXA) is the standard for bone mineral density (BMD) assessment but faces accessibility and cost limitations for widespread screening.
- Accurate BMD estimation is crucial for diagnosing and monitoring bone health conditions.
Purpose of the Study:
- To evaluate the efficacy of artificial neural networks (ANN) in estimating site-specific BMD values.
- To compare ANN-derived BMD estimates with DXA reference values using anthropometric data.
Main Methods:
- An artificial neural network (ANN) model was developed to predict BMD.
- Input variables included anthropometric measurements: sex, age, weight, height, BMI, waist-to-hip ratio, and skinfold thicknesses.
- Output BMD values (spine, pelvis, total) were compared against DXA reference data.
Main Results:
- ANN-generated BMD estimates closely matched DXA reference values across all studied groups.
- The model demonstrated high accuracy in predicting site-specific BMD using limited anthropometric inputs.
Conclusions:
- Artificial neural networks provide a viable and accurate method for estimating BMD.
- This approach holds potential for cost-effective, accessible bone health screening and prediction.