Development of a Carotid Vulnerable Plaque Phantom Model Evaluated by Pixel Distribution Analysis.

ABSTRACT

Carotid artery plaque composed of a larger percentage of lipids and/or intra-plaque hemorrhage are considered "vulnerable" or at higher risk for rupture. It is thought that such vulnerable lesions contribute to the majority of cardiovascular events. Ultrasound may facilitate the identification of plaque tissue types associated with risk for rupture. Pixel distribution analysis (PDA) is a plaque composition imaging analysis method that assigns grayscale ranges to corresponding tissue types. The aim of this study was to develop an in vitro vulnerable carotid plaque mimic (phantom) using known rat tissue types (fat, muscle and bone) to establish corresponding PDA ranges. Two sets of PDA grayscale ranges were established (i) the combined tissue set, which combined tissue subtypes into their respective categories-polyvinyl chloride (representing blood, grayscale range 0-4), muscle (84-95), fat (99-113) and bone (145-175); (ii) Individual tissue set for each tissue subtype-polyvinyl chloride (grayscale range 0-4), neck muscle (68-86), leg muscle (76-86), epididymal fat (91-100), abdomen muscle (104-108), subcutaneous fat (111-120) and bone (145-175). The grayscale pixel range overlaped between tissue types (87-90 and 109-110). These ranges were tested on five simulated polyvinyl chloride heterogeneous plaque types containing epididymal fat, leg muscle, neck muscle, abdominal muscle or bone. The individual tissue set grayscale ranges detected significantly more pixels within the correct tissue category than the combined tissue set ranges (≤10.1%, p < 0.05). This study represents a novel phantom PDA method to assess plaque heterogeneity and may be used to infer tissue type composition in clinical imaging. Additionally, this plaque phantom may serve as a platform for development and testing of novel composition analysis methods.