Related Experiment Videos
Automated volumetry at CT colonography: a phantom study
Meghan E Blake1, Jorge A Soto, Robert A Hayes
1Department of Radiology, Boston University School of Medicine, 88 East Newton Street, Boston, MA 02118, USA. mebblake@yahoo.com
Academic Radiology
|May 4, 2005
Summary
A computer-aided detection (CAD) algorithm accurately measured polyp-like structures in a phantom using CT colonography (CTC) images. The CAD software demonstrated high accuracy in determining both volume and maximum linear dimension, with minimal error.
Area of Science:
- Medical imaging
- Radiology
- Computer-aided detection
Background:
- Accurate measurement of polyp-like structures is crucial for colorectal cancer screening.
- Computer-aided detection (CAD) algorithms offer potential for improving the accuracy and efficiency of image analysis.
Purpose of the Study:
- To evaluate the performance of a CAD algorithm in quantifying polyp-like structures on CT colonography (CTC) images.
- To assess the accuracy of CAD in measuring the volume and maximum linear dimension of simulated polyps.
Main Methods:
- A custom Plexiglas phantom with idealized polyp-like objects (spheres, hemispheres) was created.
- The phantom was imaged using a four-channel detector CT scanner with varying parameters.
- A CAD algorithm was applied to segment surfaces and measure polypoid regions of interest.
Main Results:
- The CAD algorithm exhibited a mean error of 3% for volume measurements of spheres and 5% for hemispheres.
- Mean error in linear dimension measurement was approximately 2% for both shapes.
- CAD-calculated measurements showed strong correlation with actual values, independent of imaging parameters.
Conclusions:
- The CAD software accurately measures the volume and greatest linear dimension of polyp-like structures in CTC imaging.
- Measurement accuracy was not significantly impacted by phantom orientation or CT exposure factors.