Aperture optimization for emission imaging: effect of a spatially varying background

Area of Science:

• Medical Imaging
• Image Analysis
• Signal Processing

Background:

• Emission imaging systems face limitations from Poisson noise in radiation detection.
• Spatially varying backgrounds in emission imaging can obscure signals, affecting image quality.
• Accurate system assessment requires considering observer performance models.

Purpose of the Study:

• To develop a method for optimizing aperture size in emission imaging.
• To account for Poisson noise and spatially varying backgrounds.
• To evaluate system performance using model observers for detection and discrimination tasks.

Main Methods:

• The study presents a method for aperture size optimization in emission imaging.
• Performance is assessed using two model observers: the Hotelling observer and the nonprewhitening matched-filter observer.
• Tasks include Gaussian signal detection and single vs. double Gaussian signal discrimination.

Main Results:

• For signal detection, enlarging the aperture optimizes performance with a uniform background.
• An inhomogeneous background leads to an optimal aperture size naturally matched to the signal.
• The discrimination task shows a finite optimum aperture for flat backgrounds, with nonuniform backgrounds favoring finer resolution.

Conclusions:

• Optimal aperture size in emission imaging is critically dependent on background characteristics.
• The method provides a framework for optimizing imaging systems based on specific tasks and background conditions.
• Understanding aperture size effects is crucial for enhancing signal detection and discrimination in emission imaging.