Premium
Effect of varying gray‐scale resolution for detectability of bone lesions in intraoral radiographs digitized for teletransmission
Author(s) -
WENZEL ANN
Publication year - 1987
Publication title -
european journal of oral sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.802
H-Index - 93
eISSN - 1600-0722
pISSN - 0909-8836
DOI - 10.1111/j.1600-0722.1987.tb01964.x
Subject(s) - radiography , digitization , grayscale , artificial intelligence , image resolution , digital radiography , nuclear medicine , computer vision , computer science , medicine , radiology , image (mathematics)
– Visual communicaltion equipments were developed early by radiologists for the transmission of analog signals, whereas today's transmission systems are based on digitization of the radiograph. For teletransmission of dental radiographs, it is essential to assess the diagnostic accuracy of various spatial and gray‐scale resolutions. Each of 83 dry mandibles was divided into four regions. By random assignment it was decided for each region whether or not a hole should be drilled. Intraoral radiographs (3 × 4 cm) were performed of each region and interpreted. The radiographs were thereafter recorded by a video camera connected to an IBM‐PC. The personal computer held a hardware digitization card defmed at a 512 × 512 spatial resolution with the possibility of selecting varying gray‐scale resolutions. All radiographs were assessed in resolutions with 256 (8 bit), 128 (7 bit), 64 (6 bit), and 32 (5 bit) shades of gray. In no case was the original radiograph more accurate than the 512 × 512 × 8 resolution. In two of the regions evaluated, the images providing 32 shades of gray were less accurate (percentage of true positives and negatives) thao the other resolutions ( P <0.05) while in one region resolutions of 8 bit depth (256 shades of gray) provided a significantly greater accuracy than did the original radiograph and the other image resolutions ( P <0.05). This was due to fewer false negative scores and was not followed by an increase in false positive scores. It can be concluded thai a 512 × 512 spatial resolution is satisfactory for the detection of bone lesions in digitized intraoral radiographs and that 64 shades of gray provide an equally good diagnostic accuracy as do the original radiograph. This might be relevant when transmission times are of importance.