Task-based image quality assessment in radiation therapy: initial characterization and demonstration with...

Area of Science:

• Medical Physics
• Radiology
• Radiation Oncology

Background:

• Current radiation therapy (RT) image quality assessment relies on subjective evaluations or physical metrics.
• Objective, task-based image quality assessment in RT (IQA-in-RT) offers a theoretically superior alternative.
• The area under the therapeutic operating characteristic curve (AUTOC) is a key metric within IQA-in-RT.

Purpose of the Study:

• To present a practical and modular framework for implementing the IQA-in-RT theory.
• To evaluate the feasibility and utility of this framework for optimizing pre-treatment imaging in RT.
• To compare AUTOC-based optimization with traditional physical measures for RT performance.

Main Methods:

• Development of a practical, modular IQA-in-RT framework.
• Computer-simulation studies to demonstrate framework application.
• Optimization of x-ray computed tomography (CT) imaging parameters (dose, reconstruction) using AUTOC.
• Analysis of randomness and bias affecting AUTOC determination.

Main Results:

• The proposed IQA-in-RT framework successfully demonstrated feasibility and utility in simulations.
• AUTOC-based optimization identified different parameters compared to physical measures.
• These differences suggest potential improvements in RT performance.
• Analysis identified sources of randomness and bias impacting AUTOC calculation.

Conclusions:

• The presented framework provides a practical solution for task-based IQA-in-RT.
• This work advances the application of AUTOC for improving RT clinical practice.
• Optimizing imaging components using AUTOC holds promise for enhancing cancer patient care.