Evaluating image quality on post-mortem cardiac CT using an anthropomorphic phantom

Background

Ischemic heart disease is a major cause of mortality worldwide. Coronary computed tomography angiography (CCTA) has been recognised as a procedure for diagnosing atherosclerotic plaques. One method used to test the diagnostic accuracy of new technical developments on the CT scanner is post-mortem imaging. This study aimed to compare image quality of CCTA on post-mortem hearts scanned inside an anthropomorphic phantom versus scanning directly on the scanner bed, and evaluate which image was most comparable to scanning in-vivo (living patients).

Methods

Ten post-mortem hearts were scanned using the two methods and ten CCTA in-vivo were included for comparison. Region of interest (ROI) measurements in both the right and left ventricles of the hearts were made and coronary vessel analysis measured plaque burden and composition. To examine the difference between each scanning method, we compared the mean and standard deviation of these measurements. The difference in image quality was also examined visually through images and a dot plot.

Results

A Wilcoxon Signed Rank test showed that ROI measurements from the two methods were significantly different. Mann-Whitney-U tests showed a significant difference between the in-vivo measurements and the two post-mortem scanning methods. Wilcoxon signed-rank tests indicated a significant difference for 4 out of 5 plaque measurements. Visually, a noisier image was seen using the phantom, though it was closer to in-vivo imaging and had a clearer plaque visualisation.

Conclusion

A significant difference in image quality between scans taken with the heart directly on the scanner bed compared to inside the phantom, with those inside the phantom being more comparable to in-vivo scans. This highlights the importance of using an appropriate scanning technique when imaging post-mortem organs.