Logistic regression equation to screen for vertical root fractures using periapical radiographs

Area of Science:

• Endodontics
• Dental Radiology
• Oral Surgery

Background:

• Vertical root fractures (VRFs) are a significant cause of tooth loss in endodontically treated teeth.
• Accurate diagnosis of VRFs can be challenging, often requiring advanced imaging techniques.
• Periapical radiographs are routinely used, but their ability to detect VRFs is limited.

Purpose of the Study:

• To create a predictive equation for screening vertical root fractures (VRFs).
• To numerically evaluate the shape characteristics of radiolucent areas on periapical radiographs of endodontically treated maxillary teeth.
• To assess the diagnostic potential of radiolucent area morphology in identifying VRFs.

Main Methods:

• Analysis of 41 periapical radiographs from endodontically treated maxillary incisors and premolars with apical radiolucencies.
• Quantification of radiolucent area shapes using "Complexity" and "Radial SD" indices by six examiners.
• Development of a predictive equation using multiple logistic regression and ROC curve analysis.

Main Results:

• The VRF group exhibited significantly higher "Complexity" and "Radial SD" values compared to the non-VRF group (P<0.05).
• The predictive equation achieved a sensitivity of 0.68, specificity of 0.80, and efficiency of 0.75.
• The logistic regression model effectively differentiated between fractured and non-fractured roots based on radiolucent area morphology.

Conclusions:

• Teeth with VRFs present more complex radiolucent areas on periapical radiographs than those without.
• A logistic regression equation derived from evaluating radiolucent area shapes can aid in screening for VRFs.
• This quantitative approach offers a valuable tool to support the clinical diagnosis of vertical root fractures.