Related Experiment Video
Updated: Jul 6, 2025
09:26
Combining Augmented Reality and 3D Printing to Display Patient Models on a Smartphone
Published on: January 2, 2020
18.3K
Application of augmented reality using automatic markerless registration for facial plastic and reconstructive
Young Chul Kim1, Chan-Ung Park2, Seok Joon Lee1
1Department of Plastic and Reconstructive Surgery, Ulsan University College of Medicine, Asan Medical Center, Seoul, South Korea.
Summary
A novel markerless augmented reality (AR) system accurately visualizes facial anatomy for surgery. This system, using machine learning, achieves 1-2 mm registration errors, aiding reconstructive procedures.
Area of Science:
- Medical Imaging
- Computer-Aided Surgery
- Augmented Reality
Background:
- Facial plastic and reconstructive surgery requires precise visualization of anatomical structures.
- Current methods may lack real-time intraoperative guidance.
- Markerless augmented reality (AR) offers a potential solution for enhanced surgical navigation.
Purpose of the Study:
- To present a novel markerless AR system for visualizing the facial region.
- To provide an intraoperative guide for facial plastic and reconstructive surgeries using machine learning-based automatic registration.
- To evaluate the accuracy of the AR system in visualizing computed tomographic (CT) data.
Main Methods:
- Developed a markerless AR system integrating machine learning for automatic registration.
- Utilized 3D cameras to capture point clouds and CT data for 3D visualization.
- Employed a two-stage registration process: preliminary (landmark detection) and precise (Iterative Closest Point algorithms).
- Calculated AR registration error as cloud-to-cloud distance in different facial territories (upper, middle, lower).
Main Results:
- The AR system achieved a mean overall registration error of 1-2 mm.
- Mean registration errors were 1.490 ± 0.384 mm (awake) and 1.948 ± 0.638 mm (intubated).
- A significant increase in lower face registration error was observed in orally intubated patients (2.325 ± 0.971 mm) compared to awake patients (1.502 ± 0.480 mm).
Conclusions:
- The markerless AR system demonstrates high accuracy for facial visualization in reconstructive surgery.
- The system provides a reliable intraoperative guide with minimal registration errors.
- Oral intubation can introduce slight inaccuracies, particularly in the lower facial region, necessitating careful consideration during surgery.
