SU-E-J-98: 3D Tracking of Interfraction Patient Setup Uncertainties Using Multiple Kinect Sensors

Area of Science:

• Medical Physics
• Biomedical Engineering
• Radiotherapy Technology

Background:

• Accurate patient positioning is crucial in radiotherapy to minimize radiation dose to healthy tissues.
• Current methods for monitoring patient setup may involve additional imaging dose.
• Understanding patient-specific anatomical variations is key to improving treatment precision.

Purpose of the Study:

• To develop and evaluate an on-board optical 3D imaging system for real-time measurement of daily patient setup uncertainties in head and neck cancer radiotherapy.
• To hypothesize that coupling a patient-specific biomechanical model with 3D surface imaging can quantify tumor and organ deformation due to misalignment.

Main Methods:

• Utilized a markerless face recognition system with 3D cameras to capture external patient anatomy.
• Developed a real-time 3D anatomical representation by integrating facial contours.
• Implemented a non-rigid iterative closest point registration algorithm using Point Feature Histogram (PFH) descriptors for surface alignment.

Main Results:

• The proposed 3D tracking system successfully monitored patient surface setup uncertainty.
• The system demonstrated the ability to track internal anatomy when integrated with a patient-specific biomechanical head and neck model.

Conclusions:

• A novel 3D head and neck tracking system was developed to monitor interfraction patient setup uncertainties in head and neck cancer patients.
• The alignment process proved effective with and without head immobilization systems.
• Future work involves coupling surface data with motion vectors to guide biomechanical model alignment.