Program Information
Evaluation of Multiple Practical Approaches of Modulated Transfer Function Measurement in Computed Tomography
S Zhang*, S Jambawalikar , T Lin , M Liu , B Peng , R Al-Senan , Y Liang , Columbia University New York Presbyterian Hospital, New York, NY
Presentations
TH-CD-601-6 (Thursday, August 3, 2017) 10:00 AM - 12:00 PM Room: 601
Purpose: To compare MTF of a CT system determined using different techniques.
Methods: ACR and GE QA phantom images were acquired on a GE Discovery CT750 HD CT scanner using routine adult head and abdomen protocols at CUMC; four consecutive scans were performed for each protocol. In-plane resolution of these images were evaluated using the following MTF measurement methods: (1) edge method proposed by Friedman et al., (2) edge method proposed by Richard et al., (3) modulation (M) method proposed by Droege et al. and (4) amplitude method modified based on the study by Staude et al. MTF calculations were also performed on the images reconstructed at different slice thicknesses and with different algorithms based on filter back projection only. MTF curves obtained from different methods were compared.
Results: For standard algorithm, MTFs determined using M method were similar to the MTFs produced from an edge spread function before reaching the limiting resolution (10%MTF). MTFs determined using amplitude method tends to be slightly higher at the low to medium range of spatial frequency (~1 to 4.5lp/cm). For images reconstructed with bone plus algorithm, MTFs determined from each method differed drastically.
Conclusion: For standard algorithm, different techniques tend to provide a similar MTF estimate and the change of slice thickness (from 5mm to 1.25mm) does not have a significant impact to the MTF estimate. However, for bone plus algorithm, due to the edge enhancement and the high noise level as well as the phantom used, the MTF estimated from different techniques varies drastically; the limiting resolution estimated by these techniques are higher than the visual assessment, except for the amplitude method done with ACR phantom. To assess the spatial resolution of a CT scanner, one needs to select a technique and phantom suitable for the clinical protocols being investigated.
Contact Email: