A Wiener filter for nuclear medicine images

Area of Science:

• Medical Imaging
• Digital Signal Processing
• Nuclear Medicine

Background:

• Digital nuclear medicine images often suffer from noise, impacting diagnostic accuracy.
• Image quality is crucial for effective diagnosis and treatment planning.

Purpose of the Study:

• To develop and implement an improved Wiener restoration filter for digital nuclear medicine images.
• To achieve significant noise suppression and detail enhancement in clinical images.

Main Methods:

• Developed a novel Wiener filter implementation optimizing mean-square error.
• Characterized object and noise power spectrums for nuclear medicine images.
• Utilized an array processor to ensure clinically acceptable processing times.

Main Results:

• The noise power spectrum for Poisson noise was determined to be constant.
• A method for estimating the object power spectrum was established.
• Clinical images processed with the filter showed pronounced noise reduction and enhanced details.

Conclusions:

• The new Wiener filter implementation effectively enhances digital nuclear medicine image quality.
• The filter offers a practical solution for improving diagnostic accuracy in nuclear medicine.