Comparative analysis of respiratory motion tracking using Microsoft Kinect v2 sensor.

Area of Science:

• Medical imaging
• Radiotherapy
• Motion tracking

Background:

• Accurate respiratory motion tracking is crucial for effective 4D CT (4DCT) and radiotherapy.
• Existing systems like Varian RPM and Anzai belts are commonly used but can be costly or cumbersome.

Purpose of the Study:

• To present and evaluate a novel, straightforward implementation of marker-less respiratory motion tracking.
• To utilize the Kinect v2 camera as a gating tool for 4DCT and radiotherapy treatments.

Main Methods:

• Developed a C# code to track respiratory motion using the Kinect v2 depth sensor.
• Tracked depth value changes at user-selected pixels on the subject's chest/abdomen.
• Compared Kinect-derived respiratory traces with Varian RPM and Anzai belt systems using two subjects.

Main Results:

• Kinect-based depth value analysis correlated well with RPM and Anzai systems for phase and amplitude binning.
• Achieved small Interquartile Range (IQR) time spans (0.077s for Subject 1, 0.164s for Subject 2) compared to established systems.
• These time spans correspond to sub-millimeter couch movement, ensuring minimal impact on 4DCT scans.

Conclusions:

• Marker-less respiratory motion tracking is feasible using the Kinect v2 camera by analyzing depth pixel value changes.
• The Kinect v2 system offers comparable accuracy to Varian RPM and Anzai belt systems.
• This implementation provides a cost-effective and accessible gating solution for radiotherapy.