Deep Learning-based Attenuation Correction for Whole-body PET - a Multi-tracer Study with F-FDG, Ga-DOTATATE, and F-Fluciclovine

Methods: Clinical whole-body PET/CT datasets of F-FDG (N = 113),  Ga-DOTATATE (N = 76), and F-Fluciclovine (N = 90) were used to train and test tracer-specific neural networks. For each tracer, forty subjects were used to train the neural network to predict attenuation maps (µ-DL). µ-DL and µ-MLAA were compared to the gold-standard µ-CT. PET images reconstructed using the OSEM algorithm with µ-DL (OSEM) and µ-MLAA (OSEM) were compared to the CT-based reconstruction (OSEM). Tumor regions of interest were segmented by two radiologists and tumor SUV and volume measures were reported, as well as evaluation using conventional image analysis metrics.

Results: µ-DL yielded high resolution and fine detail recovery of the attenuation map, which was superior in quality as compared to µ-MLAA in all metrics for all tracers. Using OSEM as the gold-standard, OSEM provided more accurate tumor quantification than OSEM for all three tracers, e.g., error in SUV for OSEM vs. OSEM: - 3.6 ± 4.4% vs. - 1.7 ± 4.5% for F-FDG (N = 152), - 4.3 ± 5.1% vs. 0.4 ± 2.8% for  Ga-DOTATATE (N = 70), and - 7.3 ± 2.9% vs. - 2.8 ± 2.3% for F-Fluciclovine (N = 44). OSEM also yielded more accurate tumor volume measures than OSEM, i.e., - 8.4 ± 14.5% (OSEM) vs. - 3.0 ± 15.0% for F-FDG, - 14.1 ± 19.7% vs. 1.8 ± 11.6% for  Ga-DOTATATE, and - 15.9 ± 9.1% vs. - 6.4 ± 6.4% for F-Fluciclovine.

Conclusions: The proposed framework provides accurate and robust attenuation correction for whole-body F-FDG,  Ga-DOTATATE and F-Fluciclovine in tumor SUV measures as well as tumor volume estimation. The proposed method provides clinically equivalent quality as compared to CT in attenuation correction for the three tracers.