Background: Dental caries is the most common childhood ailment and one of the most prevalent chronic diseases of people worldwide. Approximately 91% of dentate adults 20 years or older have experienced dental caries. Carious lesions on proximal surfaces are particularly difficult to detect clinically, which is why intraoral radiography has become a supplemental method used to aid diagnosis. It was hypothesized that the accuracy of detection of proximal caries with the unaided eye would be reduced in sub-optimal viewing conditions; furthermore, the use of computer-assisted diagnostic software (Logicon Caries Detector) would improve the overall accuracy of observers. Methods: Eighteen radiographs with 214 surfaces were evaluated by 12 observers (general dental practitioners acting as attending faculty from University of Louisville School of Dentistry). Each observer viewed the designated surfaces in each of the viewing conditions, both with and without LCD software. The viewing conditions included: 1) brightly lit room 2) dark room, and 3) brightly lit room with plastic infection control bags on the monitors. The sensitivity and specificity of each evaluator were calculated and compared for each of the conditions using ANOVA at the significance level of p=0.05. Results: Sensitivity was significantly worse in the protective barrier condition than in the dark and bright rooms. With sharpening alone, diagnostic ability increased up to 8.59%, and LCD increased accuracy up to 17.22%. The data for specificity was slightly scattered due to a small sample size of caries-free surfaces. Conclusions: In such conditions when poor lighting and barriers can take away from diagnostic ability, computer assisted software can be a useful tool to help dentists perform as well, or better than in optimal conditions.
【 预 览 】
附件列表
Files
Size
Format
View
Effect of viewing conditions on the detection of proximal dental caries in intraoral digital images with and without computer-assisted diagnosis.
PDF
download
878987797
028-85220240
