SU-E-J-90: 2D/3D Registration Using KV-MV Image Pairs for Higher Accuracy Image Guided Radiotherapy

Area of Science:

• Medical Physics
• Radiotherapy Technology
• Image-Guided Interventions

Background:

• Tumor motion is a major source of uncertainty in radiotherapy dose delivery, impacting treatment accuracy.
• 2D/3D registration is vital for online tumor motion tracking during radiotherapy.
• A key limitation of single-projection imaging in 2D/3D registration is the inability to resolve movement along the imaging beam axis.

Purpose of the Study:

• To investigate the impact of using paired mega-voltage (MV) and kilo-voltage (kV) images on 2D/3D registration accuracy.
• To enhance tumor motion tracking precision during radiotherapy through improved registration techniques.
• To evaluate the effectiveness of combining different 2D image modalities for motion resolution.

Main Methods:

• Compared registration accuracy using single 2D images (1-kV), paired kV-MV images (1kV-1MV), and paired kV images (2-kV).
• Evaluated registration performance using 250 initial displacement starting points for each image combination.
• Measured mean target registration error (mTRE) and success rate across four different merit functions.

Main Results:

• Using paired images (kV-MV or kV-kV) reduced the root-mean-square mTRE from 6.4 mm to 2.1 mm with the MI merit function.
• The success rate for tumor motion tracking increased from 62% to 99.6% when using image pairs.
• Paired kV-MV images demonstrated slightly lower accuracy than paired kV images but offered significant improvements over single images.

Conclusions:

• Combining kV and MV images significantly improves 2D/3D registration accuracy for tumor motion monitoring in radiotherapy.
• The kV-MV image pair enables accurate resolution of motion in six degrees of freedom, enhancing treatment precision.
• Simultaneous acquisition of kV-MV images is workflow-efficient and a robust method for improving registration accuracy.