Correction to Improve X-Ray Focal Spot Measurements by Image Detector Blur Deconvolution
R Rana*, A Jain, D Bednarek, S Rudin, Univ. at Buffalo (SUNY) School of Med., Buffalo, NYSU-E-I-94 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
To deconvolve the inherent blur of a standard flat panel detector (FPD) when measuring x-ray focal spots and to evaluate the method for three focal spots and different magnifications.
A 30 micron pinhole was used to get the images of three focal spots (small, medium and large) of an x-ray tube using a FPD for different magnifications (3.16,2.66 and 2.16) implemented by varying the source to image distance. The tube parameters were kept low (50kVp) to avoid spurious effects like blooming, pinhole penetration, etc. The detector blur, characterized by its MTF was divided into the Fourier Transform of the focal spot images and result inverse transformed to obtain a deconvolved focal spot spread function from which the full width at tenth maximum (FWTM) focal spot measurements were obtained and compared with those of the uncorrected original profiles.
We measured the corrected focal spot size for small, medium and large focal spot using FWTM to be 0.41mm, 0.57mm, and 1.54mm respectively. The corrected focal spot sizes demonstrated reduced values of up to 25%, 18% and .7% compared to the uncorrected for the small, medium, and large focal spots respectively. Also, we observed that the difference between the focal spot size before and after deconvolution increases with the decrease in magnification (smaller SIDs) supporting the fact that the effect of detector blur can be reduced significantly by using a higher magnification.
For improved accuracy in x-ray focal spot measurements, deconvolving the detector resolution characteristics can be used to get rid of the excessive blur which can also be minimized by using the maximum magnification possible.
Funding Support, Disclosures, and Conflict of Interest: NIH Grant 2R01EB002873