[Potential uncertainty about image registration in thoracic image-guided radiotherapy]

Area of Science:

• Radiation Oncology
• Medical Imaging
• Radiotherapy Physics

Background:

• Image-guided radiotherapy (IGRT) is crucial for accurate cancer treatment delivery.
• IGRT utilizes imaging to correct patient setup errors before each radiation fraction.
• Image registration, the process of aligning images, can introduce its own source of error.

Purpose of the Study:

• To evaluate the uncertainty associated with image registration in IGRT.
• To quantify the magnitude of image registration uncertainty in patients with thoracic esophageal cancer.
• To determine the relationship between daily setup errors and image registration uncertainty.

Main Methods:

• Retrospective analysis of 318 image registration datasets from 12 patients with thoracic esophageal cancer undergoing IGRT.
• Two radiation therapists performed daily 3D image registration between planning CT and cone-beam CT (CBCT).
• Geometric differences were assessed by two independent radiation therapists to quantify registration uncertainty.

Main Results:

• Mean 3D displacement vectors indicated setup errors ranging from 4.9 to 15.5 mm.
• Image registration uncertainty ranged from 0.7 to 2.2 mm.
• A significant positive correlation was found between setup error magnitude and image registration uncertainty (r = 0.487, p = 0.016).

Conclusions:

• IGRT effectively corrects patient setup errors, but inherent uncertainty in image registration remains.
• The accuracy of image registration in IGRT can be influenced by the magnitude of the initial setup error.
• Minimizing setup errors is important for optimizing the reliability of image registration in IGRT.