A Novel Thresholding Based Algorithm for Detection of Vertical Root Fracture in Nonendodontically Treated Premolar...

Area of Science:

• Dentistry
• Medical Imaging
• Computer-Aided Diagnosis

Background:

• Vertical root fractures (VRFs) are a significant cause of tooth loss.
• Accurate detection of VRFs is crucial for effective treatment planning.
• Conventional radiographic methods often face limitations in diagnosing VRFs.

Purpose of the Study:

• To propose an efficient algorithm for detecting VRFs in periapical (PA) and cone-beam computed tomography (CBCT) images.
• To compare the diagnostic performance of the algorithm on PA versus CBCT radiographs.
• To evaluate the algorithm's accuracy in identifying VRFs based on fracture size.

Main Methods:

• Image denoising using block matching 3-D filtering and principal component analysis.
• Adaptive thresholding algorithm based on the modified Wellner model for fracture and canal segmentation.
• Continuous wavelet transform for VRF identification with optimal sub-image selection.

Main Results:

• The algorithm achieved high specificity for both PA (99.69 ± 0.22%) and CBCT (99.02 ± 0.77%) radiographs.
• Sensitivity ranged from 61.90% to 77.39% for PA images and 79.54% to 100% for CBCT images.
• CBCT imaging demonstrated superior performance compared to PA imaging for VRF detection.

Conclusions:

• The proposed algorithm efficiently detects vertical root fractures in dental radiographs.
• CBCT imaging is a more effective modality than PA radiography for identifying VRFs.
• This algorithm shows potential for clinical application in diagnosing VRFs.