Related Experiment Videos
Automatic localization of implanted seeds from post-implant CT images
Haisong Liu1, Gang Cheng, Yan Yu
1Department of Radiation Oncology, University of Rochester, 601 Elmwood Avenue, Rochester, NY 14642, USA. Haisong_Liu@urmc.rochester.edu
Physics in Medicine and Biology
|May 27, 2003
Summary
This study presents an automatic method for localizing implanted seeds in prostate brachytherapy using computed tomography (CT) images. The technique achieves over 99% detection rate with high accuracy and speed, aiding post-implant dosimetry.
Area of Science:
- Medical Imaging
- Radiotherapy Physics
Background:
- Prostate brachytherapy involves implanting radioactive seeds for cancer treatment.
- Accurate localization of these seeds post-implantation is crucial for effective radiation dose delivery and patient safety.
Purpose of the Study:
- To develop and validate an automated method for precise localization of implanted seeds from post-implant computed tomography (CT) images.
- To assess the accuracy, speed, and reliability of the proposed localization technique.
Main Methods:
- Segmentation of bright areas (seeds) in CT slices using binary thresholding.
- Geometry-based filtering to split large areas potentially containing multiple seeds.
- Analysis of area connectivity along the longitudinal axis to combine segmented areas into single objects.
- Calculation of seed positions using the weighted centroid of identified objects.
Main Results:
- The method was successfully tested on a prostate phantom and patient CT data.
- Achieved a seed detection rate exceeding 99%.
- Demonstrated reliable localization accuracy and high processing speed.
Conclusions:
- The developed automatic localization method is reliable and convenient for implanted seeds on CT.
- This technique can significantly assist in post-implant dosimetry for prostate brachytherapy.