LONGITUDINAL IMAGE DEBLURRING IN SPIRAL CT

PURPOSE: To assess the feasibility of digital deconvolution techniques to improve longitudinal resolution of spiral computed tomography (CT) multiplanar reformations and evaluate how technical factors in deconv olution affect longitudinal resolution, noise, and edge ringing. MATER IALS AND METHODS: Longitudinal line spread function (LSF) of the syste m was estimated from longitudinal reformations of trans-axial spiral C T images of a step test phantom. By using the estimated LSF, longitudi nal reformations of the phantom and three clinical spiral CT studies w ere deconvolved by the methods of Wiener filtering and constrained ite rative deconvolution. Edge ringing and image noise were quantified for Wiener filtering and constrained iterative deconvolution. RESULTS: Lo ngitudinal reformations were substantially deblurred and resolution im proved after deconvolution. Anatomic boundaries in clinical images wer e more clearly delineated after restoration. The methods of Wiener dec onvolution and constrained iterative deconvolution improved the sharpn ess of the phantom step boundary at the expense of increased edge ring ing and image noise. CONCLUSION: In longitudinal spiral CT reformation s, blurring along the longitudinal axis can be reduced by Wiener filte ring or constrained iterative deconvolution.