Augmented Reality With Dynamic Anatomy Modelling for Knee Arthroscopy.

Area of Science:

• Medical imaging and computer-assisted surgery
• Biomechanical modeling and simulation

Background:

• Augmented reality (AR) in knee arthroscopy often suffers from inaccuracies due to unaddressed knee flexion.
• Dynamic patient anatomy changes, particularly knee movement, are critical challenges in AR surgical navigation.

Purpose of the Study:

• To develop and validate a dynamic anatomy modeling approach for AR-assisted knee arthroscopy.
• To minimize AR errors caused by knee flexion during surgical procedures.

Main Methods:

• An association model was created linking finite-element models of knee surface and bone elements.
• Real-time knee shape during flexion was captured using a color and depth camera.
• The model dynamically deformed to reflect knee flexion states, validated with CT data.

Main Results:

• The association model successfully compensated for femur and tibia movement during knee flexion.
• The model achieved an error margin of 3.85 mm in the drilling area.
• Dynamic anatomy modeling significantly improved AR accuracy in simulated surgical conditions.

Conclusions:

• The proposed dynamic anatomy model effectively addresses knee movement challenges in AR arthroscopy.
• This approach enables more precise and reliable AR-based surgical navigation for knee surgeries.
• Further integration into surgical systems could enhance patient outcomes.