Related Experiment Videos
Application of the EM algorithm to radiographic images.
J C Brailean1, D Little, M L Giger
1Department of Electrical Engineering and Computer Science, Northwestern University, Evanston, Illinois 60208.
Medical Physics
|September 1, 1992
Summary
The expectation maximization (EM) algorithm enhances radiographic images by improving signal-to-noise ratio (SNR). It outperforms other methods for small objects, making it valuable for medical imaging analysis.
Area of Science:
- Medical Imaging
- Image Processing
- Radiography
Background:
- The Expectation Maximization (EM) algorithm is widely used in Positron Emitted Tomography (PET) for image reconstruction and restoration.
- Its application in radiographic image restoration, without reconstruction, warrants investigation for enhanced diagnostic quality.
Purpose of the Study:
- To evaluate the restoration capabilities of the EM algorithm on radiographic images.
- To quantitatively assess its performance against other image enhancement techniques using a perceived signal-to-noise ratio (SNR).
Main Methods:
- The EM algorithm was applied to radiographic images for restoration.
- Image quality was measured using a perceived SNR metric incorporating visual response and internal noise.
- Performance was compared to global contrast enhancement (windowing) and unsharp mask filtering.
Main Results:
- The EM algorithm demonstrated superior performance in enhancing radiographic images.
- Its effectiveness was particularly notable for images containing objects smaller than 4 mm.
- Relative SNR improvements indicated the EM algorithm's advantage over windowing and unsharp masking.
Conclusions:
- The EM algorithm is a powerful tool for radiographic image restoration.
- It offers significant advantages over traditional methods like windowing and unsharp masking for detecting small features.
- This technique holds promise for improving diagnostic accuracy in medical imaging.