Optimization of an Adaptive SPECT System with the Scanning Linear Estimator

Area of Science:

• Medical Imaging
• Computational Imaging
• Image Reconstruction

Background:

• Adaptive imaging systems dynamically adjust hardware configurations to enhance image quality for specific tasks.
• Optimization of Single Photon Emission Computed Tomography (SPECT) systems is crucial for improving diagnostic accuracy.

Purpose of the Study:

• To present a method for optimizing an adaptive SPECT system to maximize performance in signal estimation tasks.
• To evaluate the effectiveness of the proposed optimization method using simulated data.

Main Methods:

• Simulated an adaptive SPECT system and employed a Markov-Chain Monte Carlo (MCMC) technique to generate object ensembles from scout data.
• Imaged object ensembles with various hardware configurations and calculated signal parameters using the Scanning Linear Estimator (SLE).
• Developed a Modified Dice Index (MDI) to quantify performance, combining signal overlap and activity estimation error for simultaneous optimization.

Main Results:

• The adaptive SPECT system demonstrated superior performance compared to non-adaptive systems, particularly under low photon count conditions (thousands of photons per projection).
• Statistical analysis across one hundred unique objects showed an average MDI of 0.84 for the adaptive system versus 0.65 for the non-adaptive system.
• The MDI effectively combined errors in activity, size, and location estimates into a single metric for comprehensive system optimization.

Conclusions:

• The proposed method enables effective optimization of adaptive SPECT systems for improved signal estimation.
• Adaptive SPECT systems offer significant advantages over non-adaptive systems in challenging low-count imaging scenarios.
• The MDI serves as a valuable metric for simultaneously optimizing SPECT systems across multiple performance parameters.