Seminar: Optimization of Photon-Counting DualEnergy Thoracic Imaging
- Date
- November 15, 2024
- Time
- 12:00 p.m. - 1:00 p.m. ET
- Location
- KHE 225
- Open To
- Physics students, faculty members, adjuncts, post-docs, staff
Student: Jeffrey Dhari
Supervisor: Dr. Jesse Tanguay
Abstract
Lung cancer is the leading cause of cancer death with the survival rates being low in its advanced stages, therefore the key to survival is detection at earlier stages. However, smaller nodules may be obscured by overlaying bone structures. Photon-counting x-ray detectors (PCDs) can produce dual-energy (DE) x-ray images of lung cancer in a single x-ray exposure. This study provides a first order assessment of the contrast-to-noise ratio (CNR) and a more robust frequency-based assessment using the detectability index. We quantified the dependence of both metrics on tube voltage and energy threshold in single exposure, DE, thoracic imaging with PCDs. This work also elucidates how the fundamental processes inherent in x-ray detection by PCDs contribute to CNR degradation. We modeled the DE CNR for five theoretical PCDs, ranging from an ideal PCD that detects every primary photon in the correct energy bin while rejecting all scattered radiation to a non-ideal PCD that suffers from charge-sharing and electronic noise, and detects scatter. Model predictions were compared with experimental data extracted from images acquired using a cadmium telluride (CdTe) PCD with two energy bins. The CNR imaging phantom simulated attenuation, scatter and contrast in lung nodule imaging, while the detectability phantom included anatomic clutter. The realistic model predicted overall trends observed in the experimental data. Quantum efficiency, electronic noise, charge sharing and scatter degraded CNR by ~50%. There was good agreement in the trends between the CNR and detectability. Both CNR and detectability increased modestly (i.e. <20%) when increasing the tube voltage from 90 kV to 130 kV. Optimal energy thresholds ranged from 50 keV to 70 keV across all tube voltages. To the best of our knowledge, this study represents the first systematic investigation of singleexposure DE imaging of lung nodules with PCDs.