Automated detection of bone metastatic changes using serial CT scans.

Area of Science:

• Oncology
• Medical Imaging
• Radiology

Background:

• Bone metastases are common in advanced cancer, causing significant morbidity.
• Early detection and tracking of bone metastases progression remain challenging.
• Current methods often focus on individual scans, limiting longitudinal analysis.

Purpose of the Study:

• To develop an automated method for detecting changes in bone metastases using longitudinal CT images.
• To improve the sensitivity and accuracy of bone metastases detection, especially in early stages.
• To provide a tool for tracking the progress of bone metastases over time.

Main Methods:

• A novel approach based on subtracting registered CT volumes.
• Utilized weighted-Demons registration and symmetric warping for accurate volume alignment.
• Incorporated noise minimization, Jacobian, and false positive suppressions for enhanced detection.

Main Results:

• The method detected changes in bone metastases within 3 minutes for the entire chest bone.
• Achieved a sensitivity of 92.59% with a low false positive rate (2.58% volume, 9.71 per patient).
• Identified 113 lesions (24%) missed by radiologists and flagged three initially normal patients as abnormal.

Conclusions:

• An automated method for detecting bone metastatic changes in the entire chest bone was successfully developed.
• The system's precision and speed, enabled by advanced registration and suppression techniques, assist in sensing minute metastatic changes.
• This method shows significant potential for aiding radiologists in early-stage bone metastases detection and monitoring.