AI-based open-source software for cephalometric analysis from limited FOV radiographs

Area of Science:

• Dental diagnostics
• Medical imaging analysis
• Artificial Intelligence in healthcare

Background:

• Traditional cephalometric analysis faces limitations with manual input and imaging constraints.
• Limited field of view (FOV) images often lack crucial cephalometric landmarks.
• Artificial intelligence (AI) offers innovative solutions for dental diagnostics.

Purpose of the Study:

• To develop and evaluate an open-source software utilizing AI for improved cephalometric analysis.
• To enhance the extraction of cephalometric values from limited FOV dental images.
• To overcome limitations of traditional cephalometric analysis methods.

Main Methods:

• Creation of predictive models to estimate missing cephalometric values from reduced FOV images.
• Utilized GridSearchCV and other algorithms for model construction and selection.
• Validated the software by comparing predicted values with actual measurements and statistical analysis on 25 orthodontic cases.

Main Results:

• The AI software demonstrated effectiveness in accurately predicting key cephalometric measurements.
• The software aligns closely with actual cephalometric measurements, showing potential for clinical reliability.
• Variations in predictive accuracy across different measurements indicate areas for future development.

Conclusions:

• AI-powered cephalometric analysis software represents a significant advancement in dental diagnostics.
• The software enhances diagnostic efficiency and reduces the need for additional X-rays, improving patient care.
• Further research should expand AI applications in dentistry to address clinical decision-making and minimize bias.