The application of Bayesian penalized likelihood reconstruction algorithm on ¹⁸F-PSMA PET/CT for prostate cancer detection

Purpose: This study aims to investigate the effect of the Bayesian penalized likelihood (BPL) reconstruction algorithm on improving the detection rate of prostate cancer lesions.

Methods: 40 patients (mean age 76.25±9.66 years) with pathologically proven prostate cancer were enrolled in this study, including 19 patients without surgical treatment and 11 patients with post-surgical biochemical recurrence. All patients underwent PET/CT (Discovery MI, GE Healthcare) and their image data were reconstructed using the BLP and non-BLP reconstruction algorithm in AW workstations.

Results: Compared to results from non-BPL reconstruction, the detection rates using the BPL algorithm were increased by 7.2% (n=29, P<0.001) and 16.5% (n=11, P<0.001) in prostate cancer patients without surgical treatment and in patients with biochemical recurrence, respectively. The results were statistically significant.

Conclusions: The BPL algorithm can significantly improve the lesion detection rate on ¹⁸F-PSMA PET/CT for prostate cancer, especially for patients with biochemical recurrence.