A novel algorithm successfully removes calcifications in CT angiography (CTA) examinations while maintaining the quality of the original images, according to a study presented Wednesday. Researchers said the algorithm has the potential to improve diagnostic accuracy of CTA exams when dense calcifications are present.
Patients with vascular stenosis have a buildup of plaque (fatty deposits containing calcium, cholesterol and other substances) inside the arteries that obstructs blood flow. Currently, non-invasive assessment of vessel obstruction in the presence of heavy calcification limits the diagnostic capabilities for accurate assessment of this stenosis. Just as light cannot travel through a wall to take a picture, X-rays have a difficult time traveling through dense matter such as calcium.
"This causes calcium blooming artifact on images making calcium appear larger than its physical size and making it difficult for radiologists to accurately assess luminal patency," said presenter and study co-author Emily Koons, a PhD student at the Mayo Clinic Graduate School and Mayo's CT Clinical Innovation Center in Rochester, MN.
Koons and colleagues evaluated the performance of a novel calcium removal algorithm for vascular exams performed on a clinical photon-counting CT (PCCT) system.
Patients undergoing CTA exams were scanned on a PCCT using a multi-energy mode and 120 kV. The study included patients with identifiable calcification in their CTA images. Of the 28 patients, 24 had abdominal CTA and the remaining four underwent lower extremity runoff CTA.
Researchers identified 331 plaques from the 28 patients enrolled in the study. The amount of calcification ranged from small in 148 plaques to medium in 123 and large in 60.
Algorithm Removes Calcium Blooming
Two sets of images were reconstructed: 70 keV virtual monoenergetic images (VMI) and calcium-removed 70 keV VMI. An experienced cardiovascular radiologist evaluated all cases, focusing on vessels inferior to the aortic bifurcation, including iliac, femoral, popliteal, tibioperoneal, anterior tibial, posterior tibial and peroneal arteries.
In 259 plaques (78%), the algorithm either completely removed the calcification or left only minimum residual calcification. The overall image quality score of the calcium-removed images was identical to that of the original images. In the original images, calcium blooming was observed in 182 (55%) plaques, which was fully eliminated in the calcium-removed images.
"Removing calcium from the image eliminates blooming artifact, revealing a more accurate luminal patency," said Koons. "Use of this commercial calcium removal algorithm can potentially help improve diagnosis from CTA exams with dense calcification present."
Improved accuracy of luminal stenosis estimations helps both physician and the patient, Koons noted, potentially avoiding unnecessary invasive vascular interventions.
Currently, phantom testing is underway to further evaluate the technology.
"We are also exploring how the algorithm can potentially increase radiologist confidence in the carotid and coronary arteries," Koons said.
Access the presentation, "Calcium Removal in Vascular Exams Using a Novel Calcium-Separation Algorithm in Photon-Counting-Detector CT," (W1-SSVA03-6) on demand at Meeting.RSNA.org.