Seminar: Anti-Correlated Noise Reduction in Triple-Energy Photon-Counting Angiography

Student: Kaitlyn Sims

Supervisor: Dr. Jesse Tanguay

Abstract

Interventional radiology provides real-time x-ray imaging to doctors and can be employed for various organs or vessels; it is commonly used for procedures involving the cardiovascular system such as a coronary angiogram. Single-exposure, dual-energy x-ray angiography using a photon-counting x-ray detector has been proposed as an alternative to digital subtraction angiography and kV-switching dual-energy angiography, however it is unable to suppress anatomic noise from bone and soft tissue simultaneously. Triple-energy, photon-counting angiography would overcome this limitation but could result in high quantum noise levels that compromise image quality. We adapted anti-correlated noise reduction (ACNR) from dual-energy angiography to improve iodine detectability in triple-energy imaging. Using a cadmium-telluride photon-counting x-ray detector with analog charge summing for charge-sharing correction, we acquired images of an iodine phantom – consisting of residual vessel-like structures that remains after subtracting an iodinated healthy artery from a diseased one – at 60, 80 and 100 kV. The vessel-like structures have diameters of 1 mm, 2 mm, and 3 mm and are filled with an iodine solution containing 400 mg/mL of iodine. Three energy bins were simulated by acquiring two separate exposures using previously theoretically optimization energy thresholds. Application of ACNR improved the signal-difference to noise ratio by a factor of 5.4-11.5 compared to without ACNR. The iodine detectability improved by a factor of 3.3-6 when ACNR was applied. Therefore, the results show that the noise reduction algorithm is able to improve the image quality of the iodine structures in triple-energy photon-counting angiography.