Background: Accumulation of lipids within coronary plaques is an important process in disease progression. However, gray-scale intravascular ultrasound images cannot detect plaque lipids effectively. Radiofrequency signal analysis could provide more accurate information on preclinical coronary plaques.
Methods: We analyzed 29 zones of mild atheroma in human coronary arteries acquired at autopsy. Two histologic groups, i.e., plaques with a lipid core (group L) and plaques without a lipid core (group N), were analyzed by automatic calculation of integrated backscatter. One hundred regions of interest were set on the target zone. Radiofrequency signals from a 50 MHz transducer were digitized at 240 MHz with 12-bit resolution. The intensity of integrated backscatter and its distribution within each plaque were compared between the two groups.
Results: Although the mean backscatter was similar between the groups, intraplaque variation of backscatter and backscatter in the axial direction were larger in group L than in group N (p = 0.02). Conventional intravascular ultrasound showed extremely low sensitivity for lipid detection, despite a high specificity. In contrast, a cut-off value>32 for the total variance of integrated backscatter identified lipid-containing plaque with a high sensitivity (85%) and specificity (75%).
Conclusion: Compared with conventional imaging, assessment of the intraplaque distribution of integrated backscatter is more effective for detecting lipid. As coronary atheroma progresses, its composition becomes heterogeneous and multi-layered. This radiofrequency technique can portray complex plaque histology and can detect the early stage of plaque progression.