Theranostic SPECT reconstruction for improved resolution: application to radionuclide therapy dosimetry

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Radiotherapy

Background:

• SPECT imaging suffers from limited spatial resolution, impacting dose estimates in small tissues.
• Current resolution modeling in SPECT reconstruction can amplify noise and introduce artifacts.
• Theranostic applications require accurate dose quantification, especially in small lesions.

Purpose of the Study:

• To introduce SPECTRE, a novel SPECT reconstruction method guided by diagnostic PET images.
• To demonstrate the principle of using high-resolution PET to improve low-resolution SPECT reconstruction in theranostics.
• To evaluate SPECTRE's performance in phantom and clinical studies.

Main Methods:

• Utilized the hybrid kernelised expectation maximisation (HKEM) algorithm within the STIR framework.
• Developed SPECTRE (single photon emission computed theranostic reconstruction) to integrate PET guidance into SPECT reconstruction.
• Applied and validated the method using dual isotope (68Ga/177Lu) phantom and clinical (64Cu/67Cu) studies.

Main Results:

• SPECTRE produced SPECT images with accuracy and recovery comparable to PET.
• The method demonstrated minimal introduction of artifacts and noise amplification.
• Achieved PET-equivalent resolution in SPECT reconstruction.

Conclusions:

• SPECTRE significantly improves quantitative accuracy and reduces noise in SPECT imaging.
• This method enhances SPECT radioactivity concentration measurements for more precise dosimetry in small structures.
• SPECTRE shows promise for clinical application in theranostics, pending further optimization.