Systematic review with hierarchical clustering analysis for the fractal dimension in assessment of skeletal bone mineral density using dental radiographs

• Published in: Oral radiology 2015-01, Vol.31 (1), p.1-13
• Main Authors: Leite, André Ferreira, de Souza Figueiredo, Paulo Tadeu, Caracas, Hugo, Sindeaux, Rafael, Guimarães, Ana Tereza Bittencourt, Lazarte, Leonardo, de Paula, Ana Patrícia, de Melo, Nilce Santos
• Format: Article
• Language: English
• Subjects: Dentistry; Imaging; Medicine; Oral and Maxillofacial Surgery; Radiology; Review Article
• ISSN: 0911-6028; 1613-9674
• DOI: 10.1007/s11282-014-0188-y

Objectives To systematically review the relationship between the fractal dimension (FD) on jaw bones and skeletal bone mineral density (BMD), focusing on the different methodologies and results in the existing literature. Methods An electronic literature search was performed to identify articles that evaluated the relationship between the mandibular and maxillary FD and BMD up to February 2014. After selecting the articles, a hierarchical clustering analysis was performed to verify the dissimilarity of the studies. The following variables were grouped into clusters, and assessed for reliability of the cluster distribution: kind of study, method for calculating FD, relationship between FD and low BMD, digital imaging, and kind of dental radiography. The relative frequencies of the categories of variables were calculated and compared. Results Only 15 studies were selected and three clusters were obtained. Most articles were in the third cluster, and primarily consisted of clinical studies that used the box-counting method to calculate FD on panoramic radiographs. A significant association was found for the variable relationship between FD and low BMD and the clusters. Most of the clinical studies were performed on trabecular bone, but three studies found differences in FD between individuals with normal mandibular cortex and individuals with cortical porosity. Conclusions The kind of study, method for calculating FD, relationship between FD and low BMD, and imaging modality should be better standardized to minimize the controversies found in published studies. The performance of FD on the cortical and trabecular bone for osteoporosis screening deserves further investigation.