Hybrid [18F]-F-DOPA PET/MRI Interpretation Criteria and Scores for Glioma Follow-up After Radiotherapy.

OBJECTIVE: 18F­fluoro-L­3,4­dihydroxyphenylalanine positron emission tomography (F­DOPA PET) is used in glioma follow-up after radiotherapy to discriminate treatment-related changes (TRC) from tumor progression (TP). We compared the performances of a combined PET and MRI analysis with F­DOPA current standard of interpretation. METHODS: We included 76 consecutive patients showing at least one gadolinium-enhanced lesion on the T1­w MRI sequence (T1G). Two nuclear medicine physicians blindly analyzed PET/MRI images. In addition to the conventional PET analysis, they looked for F­DOPA uptake(s) outside T1G-enhanced areas (T1G/PET), in the white matter (WM/PET), for T1G-enhanced lesion(s) without sufficiently concordant F­DOPA uptake (T1G+/PET), and F­DOPA uptake(s) away from hemorrhagic changes as shown with a susceptibility weighted imaging sequence (SWI/PET). We measured lesions' F­DOPA uptake ratio using healthy brain background (TBR) and striatum (T/S) as references, and lesions' perfusion with arterial spin labelling cerebral blood flow maps (rCBF). Scores were determined by logistic regression. RESULTS: 53 and 23 patients were diagnosed with TP and TRC, respectively. The accuracies were 74% for T/S, 76% for TBR, and 84% for rCBF, with best cut-off values of 1.3, 3.7 and 1.25, respectively. For hybrid variables, best accuracies were obtained with conventional analysis (82%), T1G+/PET (82%) and SWI/PET (81%). T1G+/PET, SWI/PET and rCBF ≥ 1.25 were selected to construct a 3-point score. It outperformed conventional analysis and rCBF with an AUC of 0.94 and an accuracy of 87%. CONCLUSIONS: Our scoring approach combining F­DOPA PET and MRI provided better accuracy than conventional PET analyses for distinguishing TP from TRC in our patients after radiation therapy.

Pulmonary Vein Varix Mimicking Prostate Cancer Metastasis on 68Ga-Prostate Specific Membrane Antigen-11 PET/CT.

The use of PET/CT with prostate-specific membrane antigen radioligands for staging prostate cancer patients presenting a biochemical recurrence is increasing. As a consequence, the number of reports of diagnostic pitfall of this imaging modality is also increasing. A 75-year-old man referred for a second episode of biochemical recurrence of prostate cancer presented an isolated intense prostate-specific membrane antigen-11 uptake from a right lung abnormality. This abnormality was suggestive of pulmonary vein varix on contrast-enhanced lung CT. The uptake remained stable 1 month after starting androgen deprivation therapy, which confirmed the diagnostic of pulmonary vein varix.

ZTE MR-based attenuation correction in brain FDG-PET/MR: performance in patients with cognitive impairment.

OBJECTIVE: One of the main challenges of integrated PET/MR is to achieve an accurate PET attenuation correction (AC), especially in brain acquisition. Here, we evaluated an AC method based on zero echo time (ZTE) MRI, comparing it with the single-atlas AC method and CT-based AC, set as reference. METHODS: Fifty patients (70 ± 11 years old, 28 men) underwent FDG-PET/MR examination (SIGNA PET/MR 3.0 T, GE Healthcare) as part of the investigation of suspected dementia. They all had brain computed tomography (CT), 2-point LAVA-flex MRI (for atlas-based AC), and ZTE-MRI. Two AC methods were compared with CT-based AC (CTAC): one based on a single atlas, one based on ZTE segmentation. Impact on brain metabolism was evaluated using voxel and volumes of interest-based analyses. The impact of AC was also evaluated through comparisons between two subgroups of patients extracted from the whole population: 15 patients with mild cognitive impairment and normal metabolic pattern, and 22 others with metabolic pattern suggestive of Alzheimer disease, using SPM12 software. RESULTS: ZTE-AC yielded a lower bias (3.6 ± 3.2%) than the atlas method (4.5 ± 6.1%) and lowest interindividual (4.6% versus 6.8%) and inter-regional (1.4% versus 2.6%) variabilities. Atlas-AC resulted in metabolism overestimation in cortical regions near the vertex and cerebellum underestimation. ZTE-AC yielded a moderate metabolic underestimation mainly in the occipital cortex and cerebellum. Voxel-wise comparison between the two subgroups of patients showed that significant difference clusters had a slightly smaller size but similar locations with PET images corrected with ZTE-AC compared with those corrected with CT, whereas atlas-AC images showed a notable reduction of significant voxels. CONCLUSION: ZTE-AC performed better than atlas-AC in detecting pathologic areas in suspected neurodegenerative dementia. KEY POINTS: • The ZTE-based AC improved the accuracy of the metabolism quantification in PET compared with the atlas-AC method. • The overall uptake bias was 21% lower when using ZTE-based AC compared with the atlas-AC method. • ZTE-AC performed better than atlas-AC in detecting pathologic areas in suspected neurodegenerative dementia.

Attenuation correction using 3D deep convolutional neural network for brain 18F-FDG PET/MR: Comparison with Atlas, ZTE and CT based attenuation correction.

One of the main technical challenges of PET/MRI is to achieve an accurate PET attenuation correction (AC) estimation. In current systems, AC is accomplished by generating an MRI-based surrogate computed tomography (CT) from which AC-maps are derived. Nevertheless, all techniques currently implemented in clinical routine suffer from bias. We present here a convolutional neural network (CNN) that generated AC-maps from Zero Echo Time (ZTE) MR images. Seventy patients referred to our institution for 18FDG-PET/MR exam (SIGNA PET/MR, GE Healthcare) as part of the investigation of suspected dementia, were included. 23 patients were added to the training set of the manufacturer and 47 were used for validation. Brain computed tomography (CT) scan, two-point LAVA-flex MRI (for atlas-based AC) and ZTE-MRI were available in all patients. Three AC methods were evaluated and compared to CT-based AC (CTAC): one based on a single head-atlas, one based on ZTE-segmentation and one CNN with a 3D U-net architecture to generate AC maps from ZTE MR images. Impact on brain metabolism was evaluated combining voxel and regions-of-interest based analyses with CTAC set as reference. The U-net AC method yielded the lowest bias, the lowest inter-individual and inter-regional variability compared to PET images reconstructed with ZTE and Atlas methods. The impact on brain metabolism was negligible with average errors of -0.2% in most cortical regions. These results suggest that the U-net AC is more reliable for correcting photon attenuation in brain FDG-PET/MR than atlas-AC and ZTE-AC methods.