Author: Hyejoo Kang, Andrew Keeler, Mathias Lehmann, Jason Patrick Luce, Ha Nguyen, John C. Roeske 👨🔬
Affiliation: Varian Imaging Laboratory, Department of Radiation Oncology, Stritch School of Medicine, Loyola University Chicago, Stritch School of Medicine Loyola University Chicago, Cardinal Bernardin Cancer Center, Loyola University Chicago 🌍
Purpose: Markerless tumor tracking (MTT) using dual-energy (DE) subtraction imaging is being considered as a non-invasive method to track tumors during lung cancer treatment. Prior to clinical implementation of DE-MTT, a well-established quality assurance (QA) is needed to instill confidence with physicists and clinicians. We developed a QA program that benchmarks the baseline values and variations of the soft tissue weighting factor (ws), contrast-to-noise ratios (CNR), and tracked tumor locations using a customized phantom.
Methods: Over a 3-month period, a linear accelerator equipped with an on-board imager was used to collect 7 datasets of images alternating between 60 and 120 kVp using fast-kV switching in a non-clinical mode. A customized phantom with simulated tumors ranging from 5-25 mm diameter was imaged at the gantry angle of 90°. The phantom consists of tissue-equivalent materials that mimic bone, lung, soft tissue, and tumors. Data analysis was conducted offline using in-house software that allows for ws optimization to produce DE subtracted images, CNR calculation, and tumor tracking using normalized cross-correlation.
Results: The QA procedure takes <10 minutes, and the data processing time is <2 minutes. The results show that the optimal ws ranges from 0.56-0.59, suggesting good stability across datasets. The average CNR values were 0.75±0.24, 1.60±0.18, 2.52 ±0.26, 3.42±0.32, and 3.70±0.22 for tumors with 5 mm, 10 mm, 15 mm, 20 mm, and 25 mm diameter, respectively. The average tracking errors were 1.39±1.14 mm, 0.79±0.39 mm, 0.65±0.31 mm, 0.68±0.31 mm, and 0.77±0.49 mm, respectively. As expected, the 5-mm-diameter tumors had the largest tracking variations due to the relatively small size.
Conclusion: The proposed QA program successfully established baseline values for ws of 0.58 ± 0.01, CNR variations of <7%, and tracking errors of <1mm for most tumor sizes. The stability of these values over time increases confidence in the use of DE-MTT.