Thoracic target volume delineation using various maximum-intensity projection computed tomography image sets for...

Area of Science:

• Medical Imaging
• Radiation Oncology
• Radiotherapy Physics

Background:

• Four-dimensional computed tomography (4D-CT) is essential for accounting for respiratory motion in thoracic radiotherapy target volume delineation.
• Maximum-intensity projection (MIP) images derived from 4D-CT aid in defining the internal gross tumor volume (iGTV) and tumor motion envelope.

Purpose of the Study:

• To quantify differences in automatically contoured target volumes for stereotactic body radiation therapy (SBRT) delivery.
• To compare MIP datasets generated using four distinct 4D-CT reconstruction methods: phase-binned (PB), phase-corrected phase-binned (PC-PB), amplitude-binned (AB), and cine CT.

Main Methods:

• MIP datasets were generated from 4D-CT scans of 28 patients with thoracic tumors exhibiting at least 1 cm of motion.
• Commercial radiation treatment planning systems were used for automatic contouring of iGTV and planning target volumes (PTVs) from each MIP dataset.
• Differences in iGTV and PTV volumes were analyzed across the four reconstruction methods.

Main Results:

• The iGTV derived from the full cine CT dataset (iGTVcine) was consistently the largest.
• 4D-CT derived iGTVs ranged from 83.8% to 99.1% of iGTVcine, with volume differences up to 4.20 cm³.
• The phase-corrected phase-binned (PC-PB) method showed the largest average and range of iGTV underestimation; amplitude-binned (AB) and PC-PB methods exhibited the most frequent duplicate image selection.

Conclusions:

• Certain 4D-CT sorting methods can significantly under-represent the iGTV compared to a full cine CT dataset, potentially impacting SBRT planning.
• The study recommends using MIPs generated from the full cine CT dataset to ensure comprehensive tumor coverage.
• Accurate iGTV delineation is critical for effective SBRT, and the choice of MIP generation method can influence treatment planning accuracy.