Actively triggered 4d cone-beam CT acquisition.

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Computational Imaging

Background:

• Traditional 4D cone-beam CT (CBCT) reconstruction uses a retrospective binning approach.
• This method leads to uneven projection distribution, inefficient radiation dose utilization, and potential image artifacts.
• A novel acquisition framework is needed to address these limitations.

Purpose of the Study:

• To develop and demonstrate a novel 4D CBCT acquisition framework.
• To actively trigger 2D projections based on the forward-predicted tumor position.
• To improve imaging dose efficiency and image quality in 4D CBCT.

Main Methods:

• Forward-predicted tumor position was determined using electromagnetic (EM) tracking or an external motion sensor.
• A Faraday cage was employed to mitigate EM-induced artifacts.
• The acquisition strategy was validated using an anthropomorphic lung phantom with sinusoidal motion.

Main Results:

• Motion-blurred 4D CBCTs were successfully acquired using both tumor position tracking methods.
• Reconstruction artifacts from EM tracking were reduced with new correction algorithms.
• Harmonized projection distribution across respiratory phases was achieved, with 90-145 projections per phase.
• Resulting dose per phase ranged from approximately 1.7-2.6 mGy.

Conclusions:

• This work presents the first experimental demonstration of an active 4D CBCT acquisition paradigm.
• The novel framework efficiently utilizes imaging dose by acquiring projections only when needed for reconstruction.
• This approach promises improved image quality and dose efficiency in 4D CBCT applications.