Mining of body composition from abdominal CT using AI can predict risk of CVD, fracture and death
USA: Computed tomography (CT)-based body composition measures derived from fully automated artificial intelligence (AI) tools can help predict the risk of cardiovascular disease, death, and even bone fractures, researchers state in a recent study published in Radiology.Abdominal CT exams are a cornerstone of general and abdominal subspecialty radiology practice and have body composition...
USA: Computed tomography (CT)-based body composition measures derived from fully automated artificial intelligence (AI) tools can help predict the risk of cardiovascular disease, death, and even bone fractures, researchers state in a recent study published in Radiology.
Abdominal CT exams are a cornerstone of general and abdominal subspecialty radiology practice and have body composition data, which mostly gets underutilized in routine clinical practice. Besides population-based prediction, CT-based AI body composition tools measuring abdominal fat, muscle attenuation, and abdominal aortic calcium give an exciting chance for more personalized cardiometabolic opportunistic screening, risk stratification, and prediction of adverse clinical outcomes.
Matthew H. Lee, the University of Wisconsin, Madison, and colleagues explored population and sex-specific thresholds for abdominal aortic calcium measures and muscle and abdominal fat at abdominal CT for predicting the risk of adverse cardiovascular events, death, and fractures in a retrospective single-center study.
The researchers applied fully automated algorithms for determining abdominal fat (L3 level), skeletal muscle (L3 level), and abdominal aortic calcium to non-contrast abdominal CT scans obtained from asymptomatic adults screened from 2004 to 2016. Longitudinal follow-up recorded adverse cardiovascular events (cerebrovascular events, heart failure, and myocardial infarction), subsequent death, and fragility fractures. To derive thresholds for body composition measures for achieving optimal ROC curve performance and high specificity (90%) for 10-year risks, receiver operating characteristic (ROC) curve analysis was performed.
The study yielded the following findings:
· Evaluation of a total of 9223 asymptomatic adults (mean age, 57 years) was done for a median follow-up of 9 years.
· For predicting death, aortic calcium and muscle attenuation had the highest diagnostic performance, and for muscle attenuation, the areas under the ROC curve of 0.76 for men and 0.72 for women.
·In men, sex-specific thresholds were higher than in women.
· The highest-performing markers for death risk were muscle attenuation in men (31 HU; 71% sensitivity; 72% specificity) and aortic calcium in women (Agatston score, 167; 70% sensitivity; 70% specificity).
· Ninety-percent specificity thresholds for muscle attenuation for death and fragility fracture risks were 23 HU in men and 13 HU in women.
· 90% specificity Agatston score thresholds were 1475 (men) and 735 (women) for aortic calcium and risk of death and adverse cardiovascular events.
"Sex-specific thresholds for automated abdominal CT–based body composition measures can be applied to predict the risk of fragility fractures, death, and adverse cardiovascular events," the authors concluded.
The findings support the benefits of mining CT images for various data.
Lee MH, Zea R, Garrett JW, Graffy PM, Summers RM, Pickhardt PJ. Abdominal CT Body Composition Thresholds Using Automated AI Tools for Predicting 10-year Adverse Outcomes. Radiology. 2022 Sep 27:220574. doi: 10.1148/radiol.220574. Epub ahead of print. PMID: 36165792.
Medha Baranwal joined Medical Dialogues as an Editor in 2018 for Speciality Medical Dialogues. She covers several medical specialties including Cardiac Sciences, Dentistry, Diabetes and Endo, Diagnostics, ENT, Gastroenterology, Neurosciences, and Radiology. She has completed her Bachelors in Biomedical Sciences from DU and then pursued Masters in Biotechnology from Amity University. She has a working experience of 5 years in the field of medical research writing, scientific writing, content writing, and content management. She can be contacted at email@example.com. Contact no. 011-43720751