The Linkage Between Bone Densitometry and Cardiovascular Disease.

Area of Science:

• Cardiovascular disease research
• Medical imaging analysis
• Machine learning in healthcare

Background:

• Dual-energy X-ray absorptiometry (DXA) traditionally assesses bone, fat, and muscle.
• Cardiovascular disease (CVD) negatively impacts bone health and body composition.
• Early detection of body composition changes can aid CVD diagnosis and treatment.

Purpose of the Study:

• To investigate the potential of machine learning (ML) models utilizing DXA data for early cardiovascular disease (CVD) detection.
• To identify specific DXA-measured parameters associated with CVD.
• To establish a novel ML-based approach for CVD risk assessment in the Qatari population.

Main Methods:

• Utilized state-of-the-art machine learning (ML) models to classify individuals into CVD and non-CVD groups based on DXA data.
• Employed logistic regression models, achieving approximately 80% accuracy.
• Conducted an ablation study to determine the discriminatory power of different DXA parameters.

Main Results:

• Individuals with CVD exhibited elevated bone mineral density, fat content, muscle mass, and bone surface area compared to the non-CVD group.
• Fat content and bone mineral density showed greater discriminatory power in identifying CVD compared to muscle mass and bone areas.
• The developed ML model demonstrated effectiveness in distinguishing between CVD and non-CVD groups.

Conclusions:

• DXA measurements, analyzed via ML, offer an innovative approach for the early detection of cardiovascular disease (CVD).
• Specific body composition parameters like fat content and bone mineral density are significantly associated with CVD.
• This study pioneers the use of ML and DXA for CVD assessment in the Qatari population, potentially improving diagnosis and treatment planning.