Related Experiment Videos
A pattern recognition method for the removal of streaking artifact in computed tomography
Radiology
|October 1, 1981
Summary
A novel computed tomography (CT) method effectively removes streaking artifacts using nearest-neighbor pattern recognition. This technique accurately corrects ray sums, significantly improving image quality in phantom studies.
Area of Science:
- Medical Imaging
- Image Processing
- Radiology
Background:
- Streaking artifacts degrade computed tomography (CT) image quality.
- Foreign objects within scanned regions are a common cause of these artifacts.
Purpose of the Study:
- To introduce a new artifact reduction technique for CT imaging.
- To address the challenge of removing streaking artifacts caused by foreign objects.
Main Methods:
- A nearest-neighbor pattern recognition algorithm was developed.
- The method identifies and recomputes ray sums intersecting foreign objects.
Main Results:
- The technique was tested on a skull phantom with lead-induced streaking artifacts.
- Excellent artifact removal was achieved, demonstrating the method's efficacy.
Conclusions:
- The proposed nearest-neighbor pattern recognition method is highly effective for removing CT streaking artifacts.
- This technique shows promise for enhancing diagnostic accuracy in CT scans.