Development of a new threshold for pericoronary fat attenuation based on 40 keV virtual monoenergetic images from dual-energy spectral CT

Objective To investigate the distinct properties of pericoronary adipose tissue (PCAT) on 40 keV virtual monoenergetic images (VMI) from spectral detector CT (SDCT) and compare them with conventional mixed-energy images (MEI), aiming to establish a novel threshold for the fat attenuation index (FAI) at this energy level. Methods We retrospectively analyzed 60 patients who underwent coronary angiography with SDCT. The PCAT around the right coronary artery was selected for FAI assessment. CT attenuation values (minimum, mean, median, maximum), value distribution, background noise, and signal-to-noise ratio (SNR) were compared between 40 keV VMI and MEI within the same volume of interest. Results Compared to MEI, 40 keV VMI demonstrated significantly lower minimum, mean, and median CT attenuation values (all P < 0.001), a higher maximum value (−30 vs. 54.00 [28.25-89.75] HU), and a broader distribution of CT values. Background noise was similar between the two modalities, while the SNR was significantly higher on 40 keV VMI ( P < 0.001). The difference in mean CT attenuation was solely associated with photon energy ( P < 0.001) and was independent of clinical factors. Based on these findings, we established a novel linear regression equation and proposed a new threshold range for FAI at 40 keV VMI. Conclusion The attenuation properties of PCAT are fundamentally different on 40 keV VMI compared to conventional MEI. Our findings underscore the critical need for energy-specific interpretation of FAI in spectral CT and provide a foundational step towards its clinical application by proposing a new threshold.