An improved MLC segmentation algorithm and software for step-and-shoot IMRT delivery without tongue-and-groove error

Area of Science:

• Medical Physics
• Computer Science
• Radiation Oncology

Background:

• Intensity-modulated radiation therapy (IMRT) delivery relies on multileaf collimator (MLC) segmentation.
• Existing MLC segmentation algorithms, like the static leaf sequencing (SLS) method, do not fully correct for tongue-and-groove errors.
• Tongue-and-groove errors can impact the accuracy of radiation dose delivery in IMRT.

Purpose of the Study:

• To introduce an improved MLC segmentation algorithm, SLS(NOTG), for step-and-shoot IMRT.
• To reduce treatment times by minimizing the number of segments in IMRT plans.
• To minimize tongue-and-groove errors in computed IMRT plans.

Main Methods:

• The SLS(NOTG) algorithm models MLC segmentation as a weighted minimum-cost path problem using graph theory.
• It takes intensity maps as input and outputs optimized leaf sequences without tongue-and-groove error.
• The algorithm's performance was compared against the CORVUS system on an Elekta LINAC.

Main Results:

• SLS(NOTG) computed up to 50% fewer segments compared to CORVUS 5.0 for the same intensity maps.
• Clinical verification confirmed that SLS(NOTG) plans achieve dose distributions matching CORVUS plans without tongue-and-groove error.
• SLS(NOTG) can minimize tongue-and-groove errors under constraints on the maximum number of segments.

Conclusions:

• SLS(NOTG) is an effective algorithm for improving step-and-shoot IMRT delivery.
• The algorithm successfully reduces treatment time and eliminates tongue-and-groove errors.
• SLS(NOTG) offers flexibility by allowing minimization of errors subject to segment number constraints.