Seminar: Experimental optimization of DE photon-counting angiography in realistic anatomical geometries (For students, faculty & staff only)

Who can attend: Open to Graduate students, faculty, staff, adjunct faculty, post-docs.

Student: Jamie Tien
Supervisor: Dr. Jesse Tanguay
Abstract

Complex anatomical geometries produce spatially varying x-ray attenuation that can limit the detectability of overlapping structures. In angiography, this challenge arises when bone overlaps iodinated arteries. Digital subtraction angiography (DSA) addresses this by subtracting pre- and post-contrast acquisitions to produce iodine-specific images but this is highly susceptible to motion artifacts, which can reduce artery detectability. Dual-energy (DE) photon-counting angiography is a promising alternative that generates material-suppressed images. Photon-counting detectors separate incoming x-ray photons according to their energies, enabling energy-dependent image formation and material-specific suppression. However, DE material suppression increases image quantum noise and motivates the need for noise reduction algorithms, such as anti-correlated noise reduction (ACNR) which can reduce image quantum noise while preserving spatial resolution. Although optimal imaging parameters for DE photon-counting angiography in simple phantom geometries have been identified in literature, the optimal imaging parameters that balance bone suppression and artery contrast-to-noise ratio (CNR) in realistic anatomical geometries have yet to be established. As such, the purpose of this work is to experimentally optimize the tube voltage and energy thresholds of DE photon-counting angiography while using the knee and spine phantoms as realistic test cases. Knee, thoracic spine, and lumbar spine anthropomorphic phantoms were 3D-printed with materials that replicate the x-ray attenuation properties of the corresponding anatomical structures. Each phantom will be imaged over a range of tube voltages and energy thresholds and processed with and without ACNR. Optimal parameters will be selected based on both artery CNR and residual bone visibility to identify parameters that provide effective suppression while maintaining artery visibility. This study will establish optimal imaging parameters in realistic anatomical geometries, providing a foundation for further development and evaluation of this technique in complex anatomical environments.