Refine
Has Fulltext
- yes (3)
Is part of the Bibliography
- yes (3)
Year of publication
- 2022 (3) (remove)
Document Type
- Journal article (3)
Language
- English (3)
Keywords
- 177Lu (1)
- DSB damage (1)
- Deep learning (1)
- Denoising (1)
- EANM dosage card (1)
- Monte Carlo (1)
- PET (1)
- PET/MR systems (1)
- SPECT (1)
- irradiation (1)
Institute
Aim
Recent advancements in PET technology have brought with it significant improvements in PET performance and image quality. In particular, the extension of the axial field of view of PET systems, and the introduction of semiconductor technology into the PET detector, initially for PET/MR, and more recently available long-field-of-view PET/CT systems (≥ 25 cm) have brought a step change improvement in the sensitivity of PET scanners. Given the requirement to limit paediatric doses, this increase in sensitivity is extremely welcome for the imaging of children and young people. This is even more relevant with PET/MR, where the lack of CT exposures brings further dose reduction benefits to this population. In this short article, we give some details around the benefits around new PET technology including PET/MR and its implications on the EANM paediatric dosage card.
Material and methods
Reflecting on EANM adult guidance on injected activities, and making reference to bed overlap and the concept of MBq.min bed\(^{-1}\) kg\(^{-1}\), we use published data on image quality from PET/MR systems to update the paediatric dosage card for PET/MR and extended axial field of view (≥ 25 cm) PET/CT systems. However, this communication does not cover the expansion of paediatric dosing for the half-body and total-body scanners that have recently come to market.
Results
In analogy to the existing EANM dosage card, new parameters for the EANM paediatric dosage card were developed (class B, baseline value: 10.7 MBq, minimum recommended activity 10 MBq). The recommended administered activities for the systems considered in this communication range from 11 MBq [\(^{18}\)F]FDG for a child with a weight of 3 kg to 149 MBq [\(^{18}\)F]FDG for a paediatric patient weight of 68 kg, assuming a scan of 3 min per bed position. The mean effective dose over all ages (1 year and older) is 2.85 mSv.
Conclusion
With this, recommendations for paediatric dosing are given for systems that have not been considered previously.
Purpose
As α-emitters for radiopharmaceutical therapies are administered systemically by intravenous injection, blood will be irradiated by α-particles that induce clustered DNA double-strand breaks (DSBs). Here, we investigated the induction and repair of DSB damage in peripheral blood mononuclear cells (PBMCs) as a function of the absorbed dose to the blood following internal ex vivo irradiation with [\(^{223}\)Ra]RaCl2.
Methods
Blood samples of ten volunteers were irradiated by adding [\(^{223}\)Ra]RaCl2 solution with different activity concentrations resulting in absorbed doses to the blood of 3 mGy, 25 mGy, 50 mGy and 100 mGy. PBMCs were isolated, divided in three parts and either fixed directly (d-samples) or after 4 h or 24 h culture. After immunostaining, the induced γ-H2AX α-tracks were counted. The time-dependent decrease in α-track frequency was described with a model assuming a repair rate R and a fraction of non-repairable damage Q.
Results
For 25 mGy, 50 mGy and 100 mGy, the numbers of α-tracks were significantly increased compared to baseline at all time points. Compared to the corresponding d-samples, the α-track frequency decreased significantly after 4 h and after 24 h. The repair rates R were (0.24 ± 0.05) h−1 for 25 mGy, (0.16 ± 0.04) h−1 for 50 mGy and (0.13 ± 0.02) h−1 for 100 mGy, suggesting faster repair at lower absorbed doses, while Q-values were similar.
Conclusion
The results obtained suggest that induction and repair of the DSB damage depend on the absorbed dose to the blood. Repair rates were similar to what has been observed for irradiation with low linear energy transfer.
Background
In recent years, a lot of effort has been put in the enhancement of medical imaging using artificial intelligence. However, limited patient data in combination with the unavailability of a ground truth often pose a challenge to a systematic validation of such methodologies. The goal of this work was to investigate a recently proposed method for an artificial intelligence-based generation of synthetic SPECT projections, for acceleration of the image acquisition process based on a large dataset of realistic SPECT simulations.
Methods
A database of 10,000 SPECT projection datasets of heterogeneous activity distributions of randomly placed random shapes was simulated for a clinical SPECT/CT system using the SIMIND Monte Carlo program. Synthetic projections at fixed angular increments from a set of input projections at evenly distributed angles were generated by different u-shaped convolutional neural networks (u-nets). These u-nets differed in noise realization used for the training data, number of input projections, projection angle increment, and number of training/validation datasets. Synthetic projections were generated for 500 test projection datasets for each u-net, and a quantitative analysis was performed using statistical hypothesis tests based on structural similarity index measure and normalized root-mean-squared error. Additional simulations with varying detector orbits were performed on a subset of the dataset to study the effect of the detector orbit on the performance of the methodology. For verification of the results, the u-nets were applied to Jaszczak and NEMA physical phantom data obtained on a clinical SPECT/CT system.
Results
No statistically significant differences were observed between u-nets trained with different noise realizations. In contrast, a statistically significant deterioration was found for training with a small subset (400 datasets) of the 10,000 simulated projection datasets in comparison with using a large subset (9500 datasets) for training. A good agreement between synthetic (i.e., u-net generated) and simulated projections before adding noise demonstrates a denoising effect. Finally, the physical phantom measurements show that our findings also apply for projections measured on a clinical SPECT/CT system.
Conclusion
Our study shows the large potential of u-nets for accelerating SPECT/CT imaging. In addition, our analysis numerically reveals a denoising effect when generating synthetic projections with a u-net. Clinically interesting, the methodology has proven robust against camera orbit deviations in a clinically realistic range. Lastly, we found that a small number of training samples (e.g., ~ 400 datasets) may not be sufficient for reliable generalization of the u-net.