"It is important to identify patients who have a plaque that is vulnerable to rupture," explained Pavlo Baturin, a member of Sabee Molloi's research group at UCI's Department of Radiological Sciences. "Our method identifies the structure of the plaque on a pixel-by-pixel basis using spectral CT imaging. Knowing the fractional concentration of calcium per voxel and its distribution in the vicinity of the plaque is vital when assessing vulnerability."

Four-material decomposition

A vulnerable plaque is a type of plaque that is particularly unstable and prone to rupture. Two complementary and independent biomarkers that can be used to evaluate a plaque's vulnerability are the degree of inflammation and the presence of spotty calcifications in the fibrous cap surrounding the plaque.

The technique proposed by Baturin, Molloi and their colleague Yahya Alivov uses two contrast agents: gold nanoparticles (AuNPs) and iodine. Following intravenous injection, AuNPs are known to accumulate in macrophages and, in turn, can be used to identify the plaque's inflammation status. Iodine introduced into the blood, on the other hand, is used to assess stenotic narrowing.

The researchers then use a post-reconstruction four-material decomposition approach, which quantifies the concentration of iodine, AuNPs, calcium and tissue simultaneously present in a single voxel, revealing the overall vulnerability of the plaque.

Energy-resolving detectors

The key enabling technology that makes this technique possible is a photon-counting energy-resolving detector. Such a detector can have multiple energy bins, meaning that several images can be acquired during one single CT scan and material-dependent attenuation can be exploited by decomposition algorithms.

To perform the four-material decomposition, Baturin and colleagues use an energy-resolving detector based on cadmium-zinc-telluride with five energy bins, where the energy thresholds of each bin are carefully chosen based on the attenuation spectrum of the materials to be identified.

"We simulate a polyenergetic X-ray spectrum," explained Baturin. "Since we use CT, the data acquired by our detector are represented by a sinogram, so we have five sinograms for our five energy bins. K-edge decomposition in conjunction with the maximum likelihood method is then applied to these sinogram data to measure the concentration of iodine and AuNPs. We then use a subtraction technique to extract the concentrations of the calcium and tissue background. Overall, this allows us to measure the concentrations of all four materials simultaneously present in a single voxel."

The team applied this technique to two phantoms. First, a PMMA phantom consisting of a 3 × 4 array of identical wells containing mixtures of tissue, calcium, iodine and AuNPs in different (known) ratios, and second, a chest phantom with a vulnerable plaque located in the coronary artery.

"Material identification on the small phantom was successful," said Baturin. "Our quantitative analysis on the more realistic chest phantom showed a high correlation between given and measured concentrations."

Experimental confirmation

The next goal for the UCI team is to validate the simulation results via experiments on phantoms and small animals. If successfully translated into a clinical environment, the technique has the potential to identify not only patients with vulnerable plaque but also those with coronary artery disease.

"When macrophages are identified together with the spotty calcifications the patient will be diagnosed as having a vulnerable plaque," noted Baturin. "When severe calcifications are found, the patient will be identified as a vulnerable patient with coronary artery disease."

Related articles in PMB
Spectral CT imaging of vulnerable plaque with two independent biomarkers
Pavlo Baturin et al Phys. Med. Biol. 57 4117
Photon counting spectral CT: improved material decomposition with K-edge-filtered x-rays
Polad M Shikhaliev Phys. Med. Biol. 57 1595
Photon counting spectral CT versus conventional CT: comparative evaluation for breast imaging application
Polad M Shikhaliev and Shannon G Fritz Phys. Med. Biol. 56 1905
Performance simulation of an x-ray detector for spectral CT with combined Si and Cd[Zn]Te detection layers
Christoph Herrmann et al Phys. Med. Biol. 55 7697