Incorporating clinical data and coronary CT angiography-based pericoronary adipose tissue attenuation for reliable prediction of abnormal bone mass

In this retrospective study, the medical records, coronary CT angiography (CCTA) and quantitative computed tomography scans of 361 patients were reviewed to identify predictors of abnormal bone mass. Pericoronary adipose tissue (PCAT) attenuation and a number of clinical variables were collected to develop two predictive models: model 1, based solely on clinical variables, and model 2, which combined clinical data with PCAT attenuation measurements. Although several clinical variables were significantly associated with abnormal bone mass, model 2 demonstrated superior predictive performance, with an area under the curve (AUC) of 0.959 compared to 0.920 for model 1. The higher clinical utility of model 2 was also confirmed by calibration-curve and decision-curve analysis. In conclusion, integrating CCTA-based PCAT attenuation with clinical variables enhances the accuracy of predictive models for bone mass abnormalities.