TU-FG-209-06: Quantitative Evaluation of the Temporal Performance of Clinical Fluoroscopic Imaging Systems: The Temporal Modulation Transfer Function (TMTF)

Authors

  • Richards T,

    1. Medical Physics Graduate Program, Duke University, Durham, NC
    2. Carl E. Ravin Advanced Imaging Laboratories, Durham, NC
    Search for more papers by this author
  • Mann S,

    1. Medical Physics Graduate Program, Duke University, Durham, NC
    2. Clinical Imaging Physics Group, and Carl E. Ravin Advanced Imaging Laboratories, Department of Radiology, Duke University Medical Center, Durham, North Carolina
    Search for more papers by this author
  • Samei E

    1. Medical Physics Graduate Program, Duke University, Durham, NC
    2. Clinical Imaging Physics Group, and Carl E. Ravin Advanced Imaging Laboratories, Department of Radiology, Duke University Medical Center, Durham, North Carolina
    Search for more papers by this author

Abstract

Purpose:

Measure the temporal modulation transfer function (TMTF) of clinical fluoroscopic flat panel imaging systems in order to accurately quantify their performance for temporally sensitive clinical tasks.

Methods:

Copper blades (0.76 mm thick, 6 cm radius) with precision-machined edges were manufactured and mounted on a voltage regulated DC motor apparatus. Images were acquired with the blade apparatus positioned at the center of the detector matrix and set in motion at a constant rotational velocity (0.66 Hz), thereby creating a rotating radio-opaque edge in both space and time. The spatio-temporal edge response function was analyzed using a single frame from the acquired image sequence. Image processing included semi-automatic detection of the center of rotation, rebinning the pixels into subsampled polar coordinates, and deconvolving the TMTF from the previously measured spatial MTF. The analysis returned the presampled TMTF. This method was applied to multiple fluoroscopic imaging systems.

Results:

Initial experiments measured the TMTF of a typical fluoroscopy unit (Philips Allura Xper FD20) using cine acquisition (60.7 kVp, 6 fps, 42.6 ms pulse-width). For this system and protocol, the measured TMTF closely matched an idealized sinc function corresponding to the Fourier transform of the measured rectangular x-ray pulse-width. Within the frequency range of 0 – 47 Hz, the measured TMTF and ideal sinc function were compared at 1.46 Hz intervals. The differences measured varied within [-0.0164, 0.0324] with a root-mean-square (rms) difference of 0.0201. These particular results suggest that very little degradation in temporal performance is attributable to the imaging hardware, but rather the measured performance is dominated by the acquisition protocol parameters, namely pulse-width and frames-per-second.

Conclusion:

This method provides a clinically tractable and accurate measurement of the TMTF. Ongoing experiments are investigating the impact of different fluoroscopy systems, image processing, and protocol choice on the measured TMTF.

Ancillary