Automatic pelvic bone segmentation for radiotherapy using diagnostic imaging workstation.

Area of Science:

• Medical Imaging
• Radiation Oncology
• Computational Anatomy

Background:

• Accurate contouring of pelvic bones is crucial for effective radiation therapy planning.
• Traditional manual contouring methods can be time-consuming and prone to inter-observer variability.

Purpose of the Study:

• To evaluate the effectiveness of a diagnostic imaging workstation in streamlining pelvic bone contouring procedures.
• To assess the impact of the workstation on the accuracy of pelvic bone contouring compared to manual methods.

Main Methods:

• Four pelvic bone structures were created using different methods, including automatic contouring with EZ Sketch and manual modification.
• Computed tomography (CT) images from 18 patients were used.
• Structures were compared using Hausdorff distance (HD), mean distance to agreement (MDA), Dice similarity coefficient (DSC), and Jaccard similarity coefficient (JSC).

Main Results:

• The diagnostic imaging workstation significantly reduced contouring time, with automated methods taking as little as 1 minute 7 seconds compared to 33 minutes 5 seconds for manual contouring.
• Automated contouring methods demonstrated improved accuracy, with ST_Zio achieving an average DSC of 0.91 and JSC of 0.83.
• The average HD and MDA were lowest for ST_Zio (16.64 and 0.99, respectively), indicating superior agreement.

Conclusions:

• Diagnostic imaging workstations are effective tools for enhancing both the efficiency and accuracy of pelvic bone contouring in radiation therapy planning.
• The use of advanced imaging workstations can lead to more precise treatment planning and potentially improved patient outcomes.