An evaluation of the origin of trabecular bone patterns using visual and digital image analysis

The purpose of this research project was to investigate the origin of the anatomical structures interpreted as trabecula bone on dental radiographic images.

Mandible sections were cut sagitally into halves. Trabecular bone was removed from each section in 4 stages. Following each stage, standardized radiographs were made, using CDR direct digital equipment. Trabecular bone in the resulting digital images was measured with 4 methods: (1) mean gray level; (2) the fractal dimension of the basic images; and, following morphological image processing, (3) counting the number of trabecular ends, intercepts, and segments (EIS) and (4) performing fractal analyses of the skeletonized images. Additionally, human visual interpretation of the collected images was conducted through a written examination. Repeated measures analysis of variance (ANOVA) was used to test for changes in measurements attributable to bone removal.

Repeated measures ANOVA indicated that the use of gray levels, fractal dimension, and morphologic operations quantifying using EIS or fractal analysis had similar performance and resulted in significant changes in measurements following bone removal ( P < .05). Visual differences were not always apparent between each stage of bone reduction. Radiometric and morphologic analysis showed measurable differences between stages.

These results imply that the inner trabecula, the junctional trabecula, and the actual cortical housing all contribute to some extent to the radiograph, although changes in the radiographic architecture are not always clinically detectible.