An open environment CT-US fusion for tissue segmentation during interventional guidance

Area of Science:

• Medical Imaging and Interventional Radiology
• Biomedical Engineering
• Oncology

Background:

• Therapeutic ultrasound (US) offers noninvasive applications like drug activation and tumor ablation.
• Effective guidance of US therapy necessitates fusion with imaging modalities like CT or MRI.
• Accurate tissue characterization is crucial for targeted US interventions.

Purpose of the Study:

• To develop and validate a CT/US fusion system for enhanced tissue characterization and interventional guidance.
• To evaluate the system's performance in a preclinical breast cancer model.
• To assess the accuracy of CT-based tissue segmentation and its correlation with histology.

Main Methods:

• Developed an open-source CT/US fusion system.
• Quantified system performance using calibration error, target registration error, and fiducial registration error.
• Utilized Met-1 tumors in mouse models for evaluating CT-based tissue segmentation based on Hounsfield units (HU).
• Validated CT segmentation with histological analysis.

Main Results:

• The CT/US fusion system achieved spatial accuracy of approximately 1 mm in differentiating soft tissue, bone, and fat.
• Region of interest analysis showed significant differences in HU between tumor and surrounding fat (p<0.005).
• CT-measured tumor area strongly correlated with histology (r²=0.92).
• Fat content decreased with increasing tumor size, showing a correlation (r²=0.51).

Conclusions:

• Combined CT/US imaging is a feasible method for guiding interventions.
• The system enables real-time tracking of the acoustic focus within CT volumes.
• CT/US fusion facilitates accurate tissue characterization around the acoustic focus, enhancing interventional precision.