A Higher-Order Polynomial Method for SPECT Reconstruction.

Area of Science:

• Medical Imaging
• Computational Imaging
• Nuclear Medicine

Background:

• Current single-photon emission computed tomography (SPECT) reconstruction relies on discrete models, often piecewise constant approximations.
• These discrete models introduce irreducible errors, limiting the quality of reconstructed SPECT images in clinical settings.

Purpose of the Study:

• To develop a higher-order polynomial method for SPECT reconstruction to overcome limitations of traditional discrete models.
• To improve accuracy, reduce model errors, and enhance image quality in SPECT imaging.

Main Methods:

• Representing SPECT data acquisition using an integral equation model.
• Approximating the integral equation solution with higher-order piecewise polynomials.
• Introducing novel regularizers based on radiotracer distribution knowledge for the new discrete system.

Main Results:

• The higher-order polynomial method significantly reduces model errors compared to traditional discrete models.
• Demonstrated superior noise suppression and artifact reduction in reconstructed SPECT images.
• Achieved a tenfold reduction in the coefficient of variation using a piecewise linear polynomial method.

Conclusions:

• Higher-order polynomial methods offer a substantial advancement over traditional discrete models for SPECT reconstruction.
• This approach leads to improved diagnostic accuracy and clinical utility of SPECT imaging.
• The developed method effectively addresses key bottlenecks in current SPECT image reconstruction technology.