An analytical approach to single photon emission computed tomography with the attenuation effect

Area of Science:

• Medical Imaging
• Computational Physics
• Image Reconstruction

Background:

• Transverse plane reconstruction from projection data is crucial for medical imaging.
• Accurate attenuation correction is essential for quantitative analysis in techniques like SPECT.
• Existing methods may struggle with precise attenuation compensation.

Purpose of the Study:

• To propose an analytically exact solution for the attenuated tomographic operator.
• To introduce a Regularizing Iterative Method (RIM) for improved image reconstruction.
• To evaluate RIM's effectiveness in Single Photon Emission Computed Tomography (SPECT) for attenuation correction.

Main Methods:

• Developed a Regularizing Iterative Method (RIM) incorporating a priori knowledge of object size, shape, and attenuation distribution.
• Utilized a relaxation factor for noise filtering during iterative reconstruction.
• Applied RIM to Single Photon Emission Computed Tomography (SPECT) data acquired using a rotating scintillation camera and mini-computer system.

Main Results:

• Achieved satisfactory attenuation correction in phantom studies, improving image contrast and signal-to-noise ratio.
• Demonstrated noise filtering capabilities with a small number of iterations.
• Preliminary clinical studies on liver SPECT images showed improved deep lesion detectability compared to filtered back projection.

Conclusions:

• The Regularizing Iterative Method (RIM) provides an effective and exact solution for attenuation correction in tomographic imaging.
• RIM enhances image quality and quantitative accuracy in SPECT.
• RIM shows promise for improved diagnostic performance in clinical applications.