Quantification of interobserver variability in image registration using cone beam CT for partial bladder...

Area of Science:

• Radiation Oncology
• Medical Imaging
• Cancer Treatment

Background:

• Partial bladder irradiation quality is enhanced by image guidance using lipiodol or bladder wall surface (BWS).
• Current image guidance procedures are manual, leading to potential interobserver variability.
• Assessing this variability is crucial for optimizing treatment accuracy.

Purpose of the Study:

• To evaluate and compare the interobserver variability between lipiodol and BWS for image guidance in partial bladder irradiation.
• To quantify the impact of different image guidance surrogates on treatment margins and volumes.

Main Methods:

• Seven observers prospectively registered cone beam CTs (CBCT) from 20 bladder cancer patients.
• Lipiodol and BWS were used for registering CBCT to reference images.
• Displacement values were analyzed to calculate interobserver variability and required margins.

Main Results:

• Lipiodol demonstrated superior interobserver agreement (<2 mm mean difference, <5 mm 95% limit of agreement) compared to BWS.
• Greater variability was observed in the superior-inferior direction for both surrogates.
• Lipiodol-based planning target volumes were significantly smaller (134 cm³ vs 166 cm³) than BWS-based volumes.

Conclusions:

• Lipiodol provides higher interobserver agreement for partial bladder irradiation image guidance than BWS.
• A larger margin is recommended for the superior-inferior direction due to observed variability.
• Utilizing lipiodol can lead to more precise treatment targeting and potentially smaller treatment volumes.