Evaluation of an efficient compensation method for quantitative fan-beam brain SPECT reconstruction

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Computational Imaging

Background:

• Fan-beam collimators enhance sensitivity and resolution in single photon emission computed tomography (SPECT) for small organs like the brain and breasts.
• Existing reconstruction algorithms address various degradation factors in SPECT imaging.

Purpose of the Study:

• To present a new, accurate, and efficient reconstruction approach for SPECT using fan-beam collimators and a circular orbit.
• To compensate for noise, scatter, attenuation, and spatially variant detector response.
• To reduce patient radiation dose and simplify procedures by approximating uniform attenuation for brain imaging.

Main Methods:

• Developed a novel reconstruction approach for SPECT with fan-beam collimators and circular orbit.
• Implemented a uniform attenuation approximation strategy to eliminate the need for transmission scans.
• Evaluated performance against the ordered-subset expectation-maximization (OSEM) method using Monte Carlo simulations.
• Conducted quantitative analyses including bias-variance, receiver operating characteristics (ROC), and Hotelling trace studies.

Main Results:

• The new reconstruction approach demonstrated good performance in accuracy and efficiency.
• It effectively compensated for key degradation factors in SPECT imaging.
• The uniform attenuation approximation reduced radiation dose and procedural complexity.
• Quantitative analyses showed comparable performance to OSEM with significantly reduced computation time.

Conclusions:

• The proposed reconstruction strategy offers a viable and efficient alternative for SPECT imaging with fan-beam collimators.
• It achieves comparable accuracy to established methods while substantially decreasing computational demands.
• This method holds promise for improving SPECT procedures, particularly for brain imaging, by enhancing efficiency and reducing patient dose.