[Combined scatter and attenuation correction for 201Tl myocardial perfusion SPECT using OS-EM algorithm]

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Radiopharmacology

Background:

• Scatter and attenuation are significant challenges in myocardial perfusion SPECT imaging.
• Accurate scatter correction is crucial for reliable quantitative analysis and improved diagnostic accuracy.
• Maximum Likelihood Expectation Maximization (ML-EM) algorithms offer robust image reconstruction capabilities.

Purpose of the Study:

• To compare two ML-EM based scatter correction methods (SC(T) and SC(E)) for 201Tl myocardial perfusion SPECT.
• To evaluate the impact of these methods on image quality and noise amplification.
• To assess the effectiveness of combined scatter and attenuation correction.

Main Methods:

• Investigated SC(T) (subtraction of scatter estimate) and SC(E) (addition of scatter estimate) using ML-EM.
• Employed triple energy window method with trapezoidal approximation for scatter fraction estimation.
• Utilized ordered subsets EM (OS-EM) algorithm for image reconstruction (6 iterations).
• Evaluated techniques using cylindrical and heart/thorax phantoms with 201Tl.
• Quantified noise amplification using percent root-mean-square uncertainty (%RMSU).

Main Results:

• Both SC(T) and SC(E) methods demonstrated significant scatter removal in 201Tl myocardial perfusion SPECT.
• The SC(E) method exhibited less noise amplification compared to the SC(T) method.
• No significant difference in overall image quality was observed between the two scatter correction methods.

Conclusions:

• Both SC(T) and SC(E) are effective in reducing scatter in 201Tl myocardial perfusion SPECT.
• SC(E) offers a potential advantage in noise management.
• Further investigation may be warranted to optimize clinical application and discern subtle differences in image quality.