D D Brüllmann1, B Röhrig, S L Sulayman, R Schulze. 1. Department of Oral Surgery and Oral Radiology, University Medical Center of the Johannes Gutenberg University Mainz, Augustusplatz 2, 55131 Mainz, Germany. bruellmd@mail.uni-mainz.de
Abstract
OBJECTIVES: The aim of this study is to evaluate if theoretically possible edge shifts induced by noise-suppression filters potentially occur on objects found in digital radiographs. Most manufacturers carry out noise-suppression filtering of their images before they are displayed to the user. It is not usually possible for the user to disrupt the function of the filters. The use of these filters can lead to deletion of small image structures. METHODS: K-files (ISO size 06, 08, 10 and 15) were placed in the root canals of 6 human teeth located in cadaver jaw segments. File tip positions were measured on original and filtered digital images by three observers. The file position was marked on each filtered image and compared with the unfiltered ones. RESULTS: The 5 × 5 pixel-sized median and mean filters caused the largest underestimation of measured lengths between -7.87 pixels and -10.8 pixels (-306.93 μm and -421.2 μm). Maximum standard deviation for length differences was found for the calculated position ( = gold standard) and the original unfiltered images with 13.31 pixels. The standard deviation found for the 5 × 5 mean and median filter was 7.62 pixels and 8.68 pixels. CONCLUSION: Different studies showed that noise-suppression filters can induce edge shifts in diagnostic radiographs. The high standard deviations found for length differences between the defined gold-standard and the original images indicate that theoretical edge shifts may not be clinically relevant for length measurements of endodontic files.
OBJECTIVES: The aim of this study is to evaluate if theoretically possible edge shifts induced by noise-suppression filters potentially occur on objects found in digital radiographs. Most manufacturers carry out noise-suppression filtering of their images before they are displayed to the user. It is not usually possible for the user to disrupt the function of the filters. The use of these filters can lead to deletion of small image structures. METHODS: K-files (ISO size 06, 08, 10 and 15) were placed in the root canals of 6 human teeth located in cadaver jaw segments. File tip positions were measured on original and filtered digital images by three observers. The file position was marked on each filtered image and compared with the unfiltered ones. RESULTS: The 5 × 5 pixel-sized median and mean filters caused the largest underestimation of measured lengths between -7.87 pixels and -10.8 pixels (-306.93 μm and -421.2 μm). Maximum standard deviation for length differences was found for the calculated position ( = gold standard) and the original unfiltered images with 13.31 pixels. The standard deviation found for the 5 × 5 mean and median filter was 7.62 pixels and 8.68 pixels. CONCLUSION: Different studies showed that noise-suppression filters can induce edge shifts in diagnostic radiographs. The high standard deviations found for length differences between the defined gold-standard and the original images indicate that theoretical edge shifts may not be clinically relevant for length measurements of endodontic files.