Facial landmark localization by curvature maps and profile analysis

Area of Science:

• Orthodontics
• Medical Imaging
• Biotechnology

Background:

• Accurate 3D facial landmarks are crucial for orthodontic examination, assessment, and treatment planning.
• Current methods rely on subjective direct observation and manual soft tissue image analysis.
• An objective, digital method for facial landmark measurement is needed.

Purpose of the Study:

• To evaluate and present an objective method for measuring selected facial landmarks.
• To utilize laser-scanning technology for facial curvature map and sagittal profile analysis.
• To compare digital measurements with traditional manual methods.

Main Methods:

• 15 participants underwent 3D facial scanning using the FastSCAN™ laser scanner.
• Curvature maps were generated, and mathematical analysis of the median-sagittal profile was performed.
• Digital measurements were compared to manual orthodontist measurements using Lin's concordance correlation coefficient (CCC).

Main Results:

• The FastSCAN™ system accurately identified facial landmarks like the glabella and nasal tip.
• Substantial agreement (Lin's CCC) was found between digital and manual measurements for 4 of 7 distances.
• High reliability was observed in consecutive measurements, though some discrepancies were linked to image quality.

Conclusions:

• The 3D laser-scanning method (FastSCAN™) provides reliable and accurate facial landmark identification.
• Digital measurements, such as intercanthal distance, showed consistency with manual methods.
• This technique offers a clinically valid, rapid alternative to manual measurements, pending improvements in scanning accuracy.