AI-enabled opportunistic measurement of liver steatosis in coronary artery calcium scans predicts cardiovascular events and mortality: an AI-CVD study
Bibliographic record
Abstract
Abstract Introduction About one-third of adults in the US have some grade of hepatic steatosis. Coronary artery calcium (CAC) scans contain more information than currently reported. We previously reported new artificial intelligence (AI) algorithms applied to CAC scans for opportunistic measurement of bone mineral density, cardiac chamber volumes, left ventricular mass, and other imaging biomarkers collectively referred to as AI-CVD. In this study, we investigate a new AI-CVD algorithm for opportunistic measurement of liver steatosis. Methods We applied AI-CVD to CAC scans from 5702 asymptomatic individuals (52% female, age 62±10 years) in the Multi-Ethnic Study of Atherosclerosis. Liver attenuation index (LAI) was measured using the percentage of voxels below 40 HU. We used Cox proportional hazards regression to examine the association of LAI with incident CVD and mortality over 15 years. These analyses were minimally adjusted by BMI and fully adjusted for known CVD risk factors. Results A total of 751 CVD and 1343 deaths accrued over 15 years. Mean±SD LAI in females and males was 38±15% and 43±13%, respectively. Participants in the highest vs. lowest quartile of LAI had greater incidence of CVD over 15 years: 19% (95% CI: 17%-22%) vs. 12% (10%-14%), respectively, p<0.0001). Individuals in the highest quartile of both LAI and CAC score (n = 386) experienced 37.3% (32.3%-42.7%) incidence of all CVD events over 15 years. Individuals in the highest quartile of LAI (Q4) compared to the lowest quartile (Q1) showed a higher risk of CVD (HR: 1.43, 95% CI: 1.08-1.89), stroke (HR: 1.77, 95% CI: 1.09-2.88), and all-cause mortality (HR: 1.36, 95% CI: 1.10-1.67) independently of CVD risk factors and Agatston CAC Score. Conclusion AI-enabled CT attenuation analysis of the entire liver visible in CAC scans provides opportunistic and actionable information for early detection of patients at elevated risk of CVD events and all-cause mortality. The clinical utility of incorporating LAI along with other opportunistic findings in CAC scans as part of the AI-CVD initiative to improve CVD risk prediction warrants investigation in other cohorts.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".