Pramono
Department of Radiology, Dr. Soetomo Hospital, Indonesia

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

THE REDUCTION OF METAL ARTIFACTS USING BAND PASS MEDIAN FILTER ON HEAD CT SCAN Nimas Rokhmatik Dayyana; Pramono; Khusnul Ain; Amilia Kartikasari; Riky Tri Yunardi
Journal of Vocational Health Studies Vol. 6 No. 1 (2022): July 2022 | JOURNAL OF VOCATIONAL HEALTH STUDIES
Publisher : Faculty of Vocational Studies, Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jvhs.V6.I1.2022.17-23

Abstract

Background: One of the artifacts found on the CT scan is a metal artifact. Metal artifacts are caused by metal objects present in patients’ bodies. The file of metal artifacts can cover the organ that will be evaluated, and it can be inferred with the pixel value (CT number) assessment of the tissue around the metal. Purpose: To determine the effect of the band pass median filter and find the optimal filter to reduce metal artifacts on the head CT scan. Method: A total of 43 samples of patients’s files from head CT-scan without any contrast were reconstructed using four band pass median filters and obtained R1, R2, R5, and R10 filters. Two radiology specialists were assessed for the reduction of metal artifacts using the ImageJ application. Result: Four variations of the filter affected the reduction of metal artifacts because the band pass median filter maintained a point that was close to its neighboring points and points that were different from its neighboring points by replacing the value of the pixel with the median value of the grey level of neighboring pixels. The optimal filter recommendation is the R1 filter because it has the largest SNR value (16.9773) and the smallest RMSE value (8.57501) so that the result of the image is more informative and has a diagnostic value. Conclusion: The four filter variations were affected by reducing metal artifacts. Images with substantial SNR and fractional RMSE values produced an image that was more informative and still had diagnostic value.