A population study of paracardial fat as a risk factor for cardiovascular diseases (based on the data of the Moscow experiment on the use of computer vision in radiodiagnosis)

Abstract

Aim to study the prevalence of paracardial fat as a risk factor for cardiovascular diseases in Moscow population using an automated analysis of the results of radiological examinations. Material and methods. The research was designed as descriptive, retrospective epidemiological study. The results of chest computed tomography of 113,408 patients served as the study data. The data was analyzed by AI services in an automated mode for the presence of paracardial fat and calculation of its volume. Results. The paracardial fat was detected in 66.5% of the examined persons. The proportion of men was 45.7%, women 54.3% (p0.001). The volume of paracardial fat fluctuated in the range from 1.0 to 1517.0 ml; the average value was 282.1 ml. The average volume of paracardial fat in men (326.0 ml) was significantly larger than in women (244.7 ml) in each age group. The clinically significant volume of paracardial fat (200 ml) was detected in 33,081 individuals (in 64.0% of people having this risk factor). The risk factor was clinically significant in 71.1% of men and in 57.9% of women (p0.001). Conclusion. The prevalence of paracardial fat in Moscow population was 5.97 per 1000 individuals. A clinically significant volume of paracardial fat was most often found in both sexes in the elderly (78.7%) and senile age groups (78.2%). Each 5 years of age increased the probability of this risk factor incidence by 1.282 times in general; and the risk of developing its clinical form by 2.981 times in particular.

Keywords

• paracardial fat
• cardiovascular disease
• artificial intelligence
• computed tomography
• opportunistic screening