The effect of interobserver variability on transrectal ultrasonography-based postimplant dosimetry

Area of Science:

• Medical Imaging
• Radiation Oncology
• Prostate Cancer Treatment

Background:

• Accurate delineation of prostate volume is crucial for effective brachytherapy.
• Transrectal ultrasonography (TRUS) is commonly used for imaging during and after prostate brachytherapy.
• Interobserver variability in image interpretation can affect treatment outcomes.

Purpose of the Study:

• To assess the variability between observers when contouring the prostate on postimplant transrectal ultrasonography (TRUS) images.
• To evaluate the impact of this contouring variability on key dosimetric parameters (D90 and V100).
• To compare interobserver variability in postimplant TRUS imaging with preimplant TRUS imaging.

Main Methods:

• Twenty prostate implants using 125I seeds were performed.
• Postimplant TRUS images were acquired, with >85% of seeds localized directly.
• Three experts independently contoured prostate volumes on pre- and postimplant TRUS images.

Main Results:

• Significant interobserver variability was found in prostate volume contouring on postimplant TRUS images (median SD 11%).
• This variability led to significant differences in calculated D90 (median SD 7.8%) and V100 (median SD 4.0%).
• Interobserver variability was less pronounced on preimplant TRUS images compared to postimplant images.

Conclusions:

• Significant interobserver differences exist in prostate contouring on postimplant TRUS images.
• These differences are smaller than those typically reported for postimplant CT images.
• Variability in prostate contouring significantly impacts dosimetric parameters (D90, V100), affecting implant quality assessment.