RING ARTEFACT IS INHERENT IN COMPUTED TOMOGRAPHY INVESTIGATION: A PHANTOM STUDY

Abstract

Purpose: Defective tube or miscalibration has been implicated as the causes of Computed Tomography (CT) ring artifacts, implying that images acquired with such scanner should demonstrate ‘’ring(s)’’ artefact. However, ‘’ring(s)’’ artefact is frequently seen in moderate to severe hydrocephalic images but lacking on non-hydrocephalic images acquired using same machine. Hence we investigate causes of persisting ‘’ring(s)’’ artifact demonstrable on hydrocephalic CT images.

Materials and Methods: The research included hydrocephalic and hydrocdphalous mimicking phantom CT images with ‘’ring’’ band(s) designated as inner, middle and outer band, and non-hydphalous CT image without ‘’ring’’ band divided into four quadrants - anterior right (AR), anterior left (AL), posterior right (PR), and posterior left (PL). The CT number was measured at four perpendicular points per ‘’ring’’ and randomly at four points in each quadrant. CT number measurement was carried out using DICOM image viewer version 22504.418.1.0., Region of Interest (ROI) = 0.10cm²), and with the outermost ‘’ring’’ band taken control, the relationships between CT number mean values between ‘’ring’’ band, and each quadrant was analyzed.

Results: Dunnette multiple comparison test shows a significant difference in CT numbers between the Inner, Middle and Outer ‘’ring’’ bands with mean pixel values -1.194, -3.167, and 0.007 for hydrocephalous mimicking phantom image. Substantial difference in CT numbers between the Outer and Middle ‘’ring’’ band and similarity between the Outer and Inner ‘’ring’’ band on hydrocephalic image was also found with mean pixel values of 2.444, -1.593 and 2.222. For the non-hydrocephalic images, similarity in CT numbers across the four quadrants was noted with mean pixel values of AR = 17.89, AL=18.56, PR=17.17, and PL=14.00.

Conclusion:  CT ring artefacts may be patient based, indicating onset of pathologic process and diagnostic indicator.