Dimensional error of three-dimensional images generated by different software.

Area of Science:

• Oral and Maxillofacial Surgery
• Radiology
• Biomedical Engineering

Background:

• Accurate three-dimensional (3D) reconstructions are crucial for dental diagnostics and treatment planning.
• Evaluating the precision of various imaging software and parameters is essential for reliable clinical outcomes.

Purpose of the Study:

• To assess the accuracy of 3D reconstructions derived from different computed tomography (CT) and cone beam CT (CBCT) scanners.
• To compare the impact of varying slice thicknesses and software systems on the fidelity of 3D mandibular models.

Main Methods:

• Ten human dry mandibles were subjected to CT and CBCT imaging.
• Digital data were processed using multiple software systems to generate 3D reconstructions.
• Linear measurements were performed on both dry mandibles and their 3D counterparts for comparison.

Main Results:

• Significant differences were observed between the linear distances of dry mandibles and their 3D reconstructions.
• Relative error ranged from 3.10% to 4.82% for CBCT and 3.40% to 5.92% for CT.
• Software usability significantly influences measurement accuracy, beyond algorithmic performance.

Conclusions:

• Discrepancies in 3D reconstructions were generally below 0.58 mm, suggesting minimal impact on clinical image quality.
• The choice of software and operator handling are critical factors affecting the accuracy of 3D measurements.
• CT and CBCT 3D reconstructions offer acceptable accuracy for most dental applications, provided careful software selection and usage.