Table 7.
Application of artificial intelligence in epicardial adipose tissue and perivascular adipose tissue.
| Study | Year | Methods | DSC | R | AUC (MACE prediction) | Median volume(cm3)(ML) | Median volume (cm3)(expert) | ICC |
|---|---|---|---|---|---|---|---|---|
| Commandeur et al. (80) | 2018 | CNNs | 0.823 | 0.926 | NA | 78.03 | 78.64 | NA |
| Oikonomouet al. (83) | 2019 | Random forest, FRP | NA | NA | 0.88 | NA | NA | 0.938 |
| Lin et al. (87) | 2020 | CCTA-based radiomic analysis | NA | NA | 0.87 | 88.9 | NA | NA |
CNNs, convolutional neural networks; FRP, fat radiomic profile; DSC, Dice score coefficient; MI, myocardial infarction; PCAT, peri-coronary adipose tissue; AUC, area under the curve; R, correlation; NA, not applicable; MACE, major adverse cardiac events; AI, Artificial Intelligence; ICC, intra-class correlation coefficient.