Efficient radiation treatment planning based on voxel importance

Area of Science:

• Medical Physics
• Computational Biology
• Radiotherapy

Background:

• Radiation treatment planning (RTP) is computationally intensive due to large voxel datasets.
• Many voxels contain limited information, leading to inefficiencies in optimization.
• Existing methods often require complex modifications to optimization algorithms.

Purpose of the Study:

• To develop an efficient method for reducing the computational complexity of RTP.
• To maintain or improve treatment plan quality while accelerating the planning process.
• To offer a complementary approach to existing RTP optimization techniques.

Main Methods:

• Propose a voxel importance scoring method based on a simplified optimization objective.
• Utilize importance sampling to select a representative subset of informative voxels.
• Solve a reduced-size optimization problem using the selected subset.

Main Results:

• Achieved up to 50x faster optimization times for intensity-modulated radiation therapy (IMRT).
• Maintained plan quality comparable to traditional, non-reduced methods.
• Requires only a single probing and sampling step for problem reduction.

Conclusions:

• The proposed method significantly accelerates RTP by reducing problem size, not by altering solvers.
• This approach offers a computationally efficient and effective solution for modern radiotherapy.
• It has the potential to broadly impact clinical radiation oncology workflows.