Total variation regularized expectation maximization reconstruction improves 68Ga-FAPI-04 PET/CT image quality as...

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Radiochemistry

Background:

• ⁶⁸Ga-FAPI PET/CT is a crucial imaging modality.
• Image reconstruction techniques impact diagnostic quality.
• Optimizing reconstruction is key for improved lesion detection.

Purpose of the Study:

• To evaluate the impact of Total Variation Regularized Expectation Maximization (TVR-EM) on ⁶⁸Ga-FAPI PET/CT image quality.
• To compare TVR-EM with Ordered Subset Expectation Maximization (OS-EM) reconstruction.
• To assess image quality metrics at different acquisition times.

Main Methods:

• Retrospective analysis of 24 patients' PET/CT data.
• Image reconstruction using OS-EM and TVR-EM at 2 and 3 min/bed.
• Quantitative analysis of liver SUVmean, image noise, SNR, lesion SUVmax, and TBR.
• Subjective image quality assessment by two radiologists.

Main Results:

• TVR-EM reconstruction demonstrated significantly lower image noise (P<0.05).
• Higher SNR, lesion SUVmax, and TBR were observed with TVR-EM (P<0.05).
• TVR-EM 2 min/bed achieved lesion SUVmax comparable to or exceeding OS-EM 3 min/bed, with the highest subjective score for TVR-EM 3 min/bed (3.92±0.19).

Conclusions:

• TVR-EM effectively reduces image noise and enhances key quantitative metrics (SNR, SUVmax, TBR).
• TVR-EM facilitates faster PET/CT acquisition without compromising image quality.
• TVR-EM offers a superior alternative to OS-EM for ⁶⁸Ga-FAPI PET/CT.