Penapisan Artifak Logam pada Citra CT-scan dengan Spatial Filter

Abstract

The metal artifact where are produced on CT image can covered some area near metal artifact so it obstructed to define pixel value (CT number), which in turn it is dificult to define between normal and pathological tissues on a-CT image so metal artifact are need to reduce for help a diagnose. This research’s aim was performed to reduce metal artifact on a-CT image.
The method of this research is used spatial filtering with convolution technique where it aplicated to a-CT image with metal artifact and use image processing tool. A phantom with and without metal was scanned for to get an image data, and so CT image from patient where it have a metal artifact. The image data are filtering by seven filter where are different oneanother. Mean square error (MSE) was calculated to assessment performance of filters and a subjective assessment by the practitioners of CT-scan.
Metal artifacts are generally formed from 2-3 pixels and has the shortest distance between metal artifacts beam of 2-3 pixels so that the filters used have a filter window size is 5 pixels. The results of filtered image in general has decreased in quality, and the results of assessment of respondents, the filter that has been chosen by the respondent is filter 2 is as much as 29.03% and 0,010703 of MSE value. Therefore we can conclude that the filter 2 is the best filter of the seven filters are used to reduce metal artifacts and still maintain the image quality of CT-scan on visualize organ that can still be used for diagnosis.