Using domain knowledge for robust and generalizable deep learning-based CT-free PET attenuation and scatter correction.
Ontology highlight
ABSTRACT:
SUBMITTER: Guo R
PROVIDER: S-EPMC9537165 | biostudies-literature | 2022 Oct
REPOSITORIES: biostudies-literature
Publications
Using domain knowledge for robust and generalizable deep learning-based CT-free PET attenuation and scatter correction.
Guo Rui R Xue Song S Hu Jiaxi J Sari Hasan H Mingels Clemens C Zeimpekis Konstantinos K Prenosil George G Wang Yue Y Zhang Yu Y Viscione Marco M Sznitman Raphael R Rominger Axel A Li Biao B Shi Kuangyu K
Nature communications 20221006 1
Despite the potential of deep learning (DL)-based methods in substituting CT-based PET attenuation and scatter correction for CT-free PET imaging, a critical bottleneck is their limited capability in handling large heterogeneity of tracers and scanners of PET imaging. This study employs a simple way to integrate domain knowledge in DL for CT-free PET imaging. In contrast to conventional direct DL methods, we simplify the complex problem by a domain decomposition so that the learning of anatomy-d ...[more]
Similar Datasets
Project description:PurposeAttenuation correction and scatter compensation (AC/SC) are two main steps toward quantitative PET imaging, which remain challenging in PET-only and PET/MRI systems. These can be effectively tackled via deep learning (DL) methods. However, trustworthy, and generalizable DL models commonly require well-curated, heterogeneous, and large datasets from multiple clinical centers. At the same time, owing to legal/ethical issues and privacy concerns, forming a large collective, centralized dataset poses significant challenges. In this work, we aimed to develop a DL-based model in a multicenter setting without direct sharing of data using federated learning (FL) for AC/SC of PET images.MethodsNon-attenuation/scatter corrected and CT-based attenuation/scatter corrected (CT-ASC) 18F-FDG PET images of 300 patients were enrolled in this study. The dataset consisted of 6 different centers, each with 50 patients, with scanner, image acquisition, and reconstruction protocols varying across the centers. CT-based ASC PET images served as the standard reference. All images were reviewed to include high-quality and artifact-free PET images. Both corrected and uncorrected PET images were converted to standardized uptake values (SUVs). We used a modified nested U-Net utilizing residual U-block in a U-shape architecture. We evaluated two FL models, namely sequential (FL-SQ) and parallel (FL-PL) and compared their performance with the baseline centralized (CZ) learning model wherein the data were pooled to one server, as well as center-based (CB) models where for each center the model was built and evaluated separately. Data from each center were divided to contribute to training (30 patients), validation (10 patients), and test sets (10 patients). Final evaluations and reports were performed on 60 patients (10 patients from each center).ResultsIn terms of percent SUV absolute relative error (ARE%), both FL-SQ (CI:12.21-14.81%) and FL-PL (CI:11.82-13.84%) models demonstrated excellent agreement with the centralized framework (CI:10.32-12.00%), while FL-based algorithms improved model performance by over 11% compared to CB training strategy (CI: 22.34-26.10%). Furthermore, the Mann-Whitney test between different strategies revealed no significant differences between CZ and FL-based algorithms (p-value > 0.05) in center-categorized mode. At the same time, a significant difference was observed between the different training approaches on the overall dataset (p-value < 0.05). In addition, voxel-wise comparison, with respect to reference CT-ASC, exhibited similar performance for images predicted by CZ (R2 = 0.94), FL-SQ (R2 = 0.93), and FL-PL (R2 = 0.92), while CB model achieved a far lower coefficient of determination (R2 = 0.74). Despite the strong correlations between CZ and FL-based methods compared to reference CT-ASC, a slight underestimation of predicted voxel values was observed.ConclusionDeep learning-based models provide promising results toward quantitative PET image reconstruction. Specifically, we developed two FL models and compared their performance with center-based and centralized models. The proposed FL-based models achieved higher performance compared to center-based models, comparable with centralized models. Our work provided strong empirical evidence that the FL framework can fully benefit from the generalizability and robustness of DL models used for AC/SC in PET, while obviating the need for the direct sharing of datasets between clinical imaging centers.
| S-EPMC9742659 | biostudies-literature
Project description:PurposeTo demonstrate the feasibility of CT-less attenuation and scatter correction (ASC) in the image space using deep learning for whole-body PET, with a focus on the potential benefits and pitfalls.Materials and methodsIn this retrospective study, 110 whole-body fluorodeoxyglucose (FDG) PET/CT studies acquired in 107 patients (mean age ± standard deviation, 58 years ± 18; age range, 11-92 years; 72 females) from February 2016 through January 2018 were randomly collected. A total of 37.3% (41 of 110) of the studies showed metastases, with diverse FDG PET findings throughout the whole body. A U-Net-based network was developed for directly transforming noncorrected PET (PETNC) into attenuation- and scatter-corrected PET (PETASC). Deep learning-corrected PET (PETDL) images were quantitatively evaluated by using the standardized uptake value (SUV) of the normalized root mean square error, the peak signal-to-noise ratio, and the structural similarity index, in addition to a joint histogram for statistical analysis. Qualitative reviews by radiologists revealed the potential benefits and pitfalls of this correction method.ResultsThe normalized root mean square error (0.21 ± 0.05 [mean SUV ± standard deviation]), mean peak signal-to-noise ratio (36.3 ± 3.0), mean structural similarity index (0.98 ± 0.01), and voxelwise correlation (97.62%) of PETDL demonstrated quantitatively high similarity with PETASC. Radiologist reviews revealed the overall quality of PETDL. The potential benefits of PETDL include a radiation dose reduction on follow-up scans and artifact removal in the regions with attenuation correction- and scatter correction-based artifacts. The pitfalls involve potential false-negative results due to blurring or missing lesions or false-positive results due to pseudo-low-uptake patterns.ConclusionDeep learning-based direct ASC at whole-body PET is feasible and potentially can be used to overcome the current limitations of CT-based approaches, benefiting patients who are sensitive to radiation from CT.Supplemental material is available for this article.© RSNA, 2020.
| S-EPMC8043359 | biostudies-literature
Project description:Whole-body attenuation correction (AC) is still challenging in combined PET/MR scanners. We describe Dixon-VIBE Deep Learning (DIVIDE), a deep-learning network that allows synthesizing pelvis pseudo-CT maps based only on the standard Dixon volumetric interpolated breath-hold examination (Dixon-VIBE) images currently acquired for AC in some commercial scanners. Methods: We propose a network that maps between the four 2-dimensional (2D) Dixon MR images (water, fat, in-phase, and out-of-phase) and their corresponding 2D CT image. In contrast to previous methods, we used transposed convolutions to learn the up-sampling parameters, we used whole 2D slices to provide context information, and we pretrained the network with brain images. Twenty-eight datasets obtained from 19 patients who underwent PET/CT and PET/MR examinations were used to evaluate the proposed method. We assessed the accuracy of the μ-maps and reconstructed PET images by performing voxel- and region-based analysis comparing the SUVs (in g/mL) obtained after AC using the Dixon-VIBE (PETDixon), DIVIDE (PETDIVIDE), and CT-based (PETCT) methods. Additionally, the bias in quantification was estimated in synthetic lesions defined in the prostate, rectum, pelvis, and spine. Results: Absolute mean relative change values relative to CT AC were lower than 2% on average for the DIVIDE method in every region of interest except for bone tissue, where it was lower than 4% and 6.75 times smaller than the relative change of the Dixon method. There was an excellent voxel-by-voxel correlation between PETCT and PETDIVIDE (R 2 = 0.9998, P < 0.01). The Bland-Altman plot between PETCT and PETDIVIDE showed that the average of the differences and the variability were lower (mean PETCT-PETDIVIDE SUV, 0.0003; PETCT-PETDIVIDE SD, 0.0094; 95% confidence interval, [-0.0180,0.0188]) than the average of differences between PETCT and PETDixon (mean PETCT-PETDixon SUV, 0.0006; PETCT-PETDixon SD, 0.0264; 95% confidence interval, [-0.0510,0.0524]). Statistically significant changes in PET data quantification were observed between the 2 methods in the synthetic lesions, with the largest improvement in femur and spine lesions. Conclusion: The DIVIDE method can accurately synthesize a pelvis pseudo-CT scan from standard Dixon-VIBE images, allowing for accurate AC in combined PET/MR scanners. Additionally, our implementation allows rapid pseudo-CT synthesis, making it suitable for routine applications and even allowing retrospective processing of Dixon-VIBE data.
| S-EPMC6910626 | biostudies-literature
Project description:The authors evaluate the energy dependent noise and bias properties of monoenergetic images synthesized from dual-energy CT (DECT) acquisitions. These monoenergetic images can be used to estimate attenuation coefficients at energies suitable for positron emission tomography (PET) and single-photon emission computed tomography (SPECT) imaging. This is becoming more relevant with the increased use of quantitative imaging by PET/CT and SPECT/CT scanners. There are, however, potential variations in the noise and bias of synthesized monoenergetic images as a function of energy.The authors used analytic approximations and simulations to estimate the noise and bias of synthesized monoenergetic images of water-filled cylinders with different shapes and the NURBS-based cardiac-torso (NCAT) phantom from 40 to 520 keV, the range of SPECT and PET energies. The dual-kVp spectra were based on the GE Lightspeed VCT scanner at 80 and 140 kVp with added filtration of 0.5 mm Cu. The authors evaluated strategies of noise suppression with sinogram smoothing and dose minimization with reduction of tube currents at the two kVp settings. The authors compared the impact of DECT-based attenuation correction with single-kVp CT-based attenuation correction on PET quantitation for the NCAT phantom for soft tissue and high-Z materials of bone and iodine contrast enhancement.Both analytic calculations and simulations displayed the expected minimum noise value for a synthesized monoenergetic image at an energy between the mean energies of the two spectra. In addition the authors found that the normalized coefficient of variation in the synthesized attenuation map increased with energy but reached a plateau near 160 keV, and then remained constant with increasing energy up to 511 keV and beyond. The bias was minimal, as the linear attenuation coefficients of the synthesized monoenergetic images were within 2.4% of the known true values across the entire energy range. Compared with no sinogram smoothing, sinogram smoothing can dramatically reduce noise in the DECT-derived attenuation map. Through appropriate selection of tube currents for high and low kVp scans, DECT can deliver roughly the same amount of radiation dose as that of a single kVp CT scan, but could be used for PET attenuation correction with reduced bias in contrast agent regions by a factor of ≈ 2.6 and slightly reduced RMSE for the total image.When DECT is used for attenuation correction at higher energies, there is a noise amplification that is dependent on the energy of the synthesized monoenergetic image of linear attenuation coefficients. Sinogram smoothing reduces the noise amplification in DECT-derived attenuation maps without increasing bias. With an appropriate selection of CT techniques, a DECT scan with the same radiation dose as a single CT scan can result in a PET image with improved quantitative accuracy.
| S-EPMC3869851 | biostudies-other
Project description:PET attenuation correction (AC) on systems lacking CT/transmission scanning, such as dedicated brain PET scanners and hybrid PET/MRI, is challenging. Direct AC in image-space, wherein PET images corrected for attenuation and scatter are synthesized from nonattenuation corrected PET (PET-nonAC) images in an end-to-end fashion using deep learning approaches (DLAC) is evaluated for various radiotracers used in molecular neuroimaging studies. One hundred eighty brain PET scans acquired using 18 F-FDG, 18 F-DOPA, 18 F-Flortaucipir (targeting tau pathology), and 18 F-Flutemetamol (targeting amyloid pathology) radiotracers (40 + 5, training/validation + external test, subjects for each radiotracer) were included. The PET data were reconstructed using CT-based AC (CTAC) to generate reference PET-CTAC and without AC to produce PET-nonAC images. A deep convolutional neural network was trained to generate PET attenuation corrected images (PET-DLAC) from PET-nonAC. The quantitative accuracy of this approach was investigated separately for each radiotracer considering the values obtained from PET-CTAC images as reference. A segmented AC map (PET-SegAC) containing soft-tissue and background air was also included in the evaluation. Quantitative analysis of PET images demonstrated superior performance of the DLAC approach compared to SegAC technique for all tracers. Despite the relatively low quantitative bias observed when using the DLAC approach, this approach appears vulnerable to outliers, resulting in noticeable local pseudo uptake and false cold regions. Direct AC in image-space using deep learning demonstrated quantitatively acceptable performance with less than 9% absolute SUV bias for the four different investigated neuroimaging radiotracers. However, this approach is vulnerable to outliers which result in large local quantitative bias.
| S-EPMC7416024 | biostudies-literature
Project description:PurposeRespiratory motion (RM) significantly impacts image quality in thoracoabdominal PET/CT imaging. This study introduces a unified data-driven respiratory motion correction (uRMC) method, utilizing deep learning neural networks, to solve all the major issues caused by RM, i.e., PET resolution loss, attenuation correction artifacts, and PET-CT misalignment.MethodsIn a retrospective study, 737 patients underwent [18F]FDG PET/CT scans using the uMI Panorama PET/CT scanner. Ninety-nine patients, who also had respiration monitoring device (VSM), formed the validation set. The remaining data of the 638 patients were used to train neural networks used in the uRMC. The uRMC primarily consists of three key components: (1) data-driven respiratory signal extraction, (2) attenuation map generation, and (3) PET-CT alignment. SUV metrics were calculated within 906 lesions for three approaches, i.e., data-driven uRMC (proposed), VSM-based uRMC, and OSEM without motion correction (NMC). RM magnitude of major organs were estimated.ResultsuRMC enhanced diagnostic capabilities by revealing previously undetected lesions, sharpening lesion contours, increasing SUV values, and improving PET-CT alignment. Compared to NMC, uRMC showed increases of 10% and 17% in SUVmax and SUVmean across 906 lesions. Sub-group analysis showed significant SUV increases in small and medium-sized lesions with uRMC. Minor differences were found between VSM-based and data-driven uRMC methods, with the SUVmax was found statistically marginal significant or insignificant between the two methods. The study observed varied motion amplitudes in major organs, typically ranging from 10 to 20 mm.ConclusionA data-driven solution for respiratory motion in PET/CT has been developed, validated and evaluated. To the best of our knowledge, this is the first unified solution that compensates for the motion blur within PET, the attenuation mismatch artifacts caused by PET-CT misalignment, and the misalignment between PET and CT.
| S-EPMC11599311 | biostudies-literature
Project description:We evaluate the accuracy of scaling CT images for attenuation correction of PET data measured for bone. While the standard tri-linear approach has been well tested for soft tissues, the impact of CT-based attenuation correction on the accuracy of tracer uptake in bone has not been reported in detail. We measured the accuracy of attenuation coefficients of bovine femur segments and patient data using a tri-linear method applied to CT images obtained at different kVp settings. Attenuation values at 511 keV obtained with a (68)Ga/(68)Ge transmission scan were used as a reference standard. The impact of inaccurate attenuation images on PET standardized uptake values (SUVs) was then evaluated using simulated emission images and emission images from five patients with elevated levels of FDG uptake in bone at disease sites. The CT-based linear attenuation images of the bovine femur segments underestimated the true values by 2.9 ± 0.3% for cancellous bone regardless of kVp. For compact bone the underestimation ranged from 1.3% at 140 kVp to 14.1% at 80 kVp. In the patient scans at 140 kVp the underestimation was approximately 2% averaged over all bony regions. The sensitivity analysis indicated that errors in PET SUVs in bone are approximately proportional to errors in the estimated attenuation coefficients for the same regions. The variability in SUV bias also increased approximately linearly with the error in linear attenuation coefficients. These results suggest that bias in bone uptake SUVs of PET tracers ranges from 2.4% to 5.9% when using CT scans at 140 and 120 kVp for attenuation correction. Lower kVp scans have the potential for considerably more error in dense bone. This bias is present in any PET tracer with bone uptake but may be clinically insignificant for many imaging tasks. However, errors from CT-based attenuation correction methods should be carefully evaluated if quantitation of tracer uptake in bone is important.
| S-EPMC3353415 | biostudies-literature
Project description:Dedicated brain positron emission tomography (PET) devices can provide higher-resolution images with much lower doses compared to conventional whole-body PET systems, which is important to support PET neuroimaging and particularly useful for the diagnosis of neurodegenerative diseases. However, when a dedicated brain PET scanner does not come with a combined CT or transmission source, there is no direct solution for accurate attenuation and scatter correction, both of which are critical for quantitative PET. To address this problem, we propose joint attenuation and scatter correction (ASC) in image space for non-corrected PET (PETNC) using deep convolutional neural networks (DCNNs). This approach is a one-step process, distinct from conventional methods that rely on generating attenuation maps first that are then applied to iterative scatter simulation in sinogram space. For training and validation, time-of-flight PET/MR scans and additional helical CTs were performed for 35 subjects (25/10 split for training and test dataset). A DCNN model was proposed and trained to convert PETNC to DCNN-based ASC PET (PETDCNN) directly in image space. For quantitative evaluation, uptake differences between PETDCNN and reference CT-based ASC PET (PETCT-ASC) were computed for 116 automated anatomical labels (AALs) across 10 test subjects (1160 regions in total). MR-based ASC PET (PETMR-ASC), a current clinical protocol in PET/MR imaging, was another reference for comparison. Statistical significance was assessed using a paired t test. The performance of PETDCNN was comparable to that of PETMR-ASC, in comparison to reference PETCT-ASC. The mean SUV differences (mean ± SD) from PETCT-ASC were 4.0% ± 15.4% (P < 0.001) and -4.2% ± 4.3% (P < 0.001) for PETDCNN and PETMR-ASC, respectively. The overall larger variation of PETDCNN (15.4%) was prone to the subject with the highest mean difference (48.5% ± 10.4%). The mean difference of PETDCNN excluding the subject was substantially improved to -0.8% ± 5.2% (P < 0.001), which was lower than that of PETMR-ASC (-5.07% ± 3.60%, P < 0.001). In conclusion, we demonstrated the feasibility of directly producing PET images corrected for attenuation and scatter using a DCNN (PETDCNN) from PETNC in image space without requiring conventional attenuation map generation and time-consuming scatter correction. Additionally, our DCNN-based method provides a possible alternative to MR-ASC for simultaneous PET/MRI.
| S-EPMC6449185 | biostudies-literature
Project description: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.
| S-EPMC6779234 | biostudies-literature
Project description:OBJECTIVE: To compare the effect of implanted medical materials on (18)F-fludeoxyglucose ((18)F-FDG) positron emission tomography (PET)/MRI using a Dixon-based segmentation method for MRI-based attenuation correction (MRAC), PET/CT and CT-based attenuation-corrected PET (PETCTAC). METHODS: 12 patients (8 males and 4 females; age 58±11 years) with implanted medical materials prospectively underwent whole-body (18)F-FDG PET/CT and PET/MRI. CT, MRI and MRAC maps as well as PETCTAC and PETMRAC images were reviewed for the presence of artefacts. Their morphology and effect on the estimation of the (18)F-FDG uptake (no effect, underestimation, overestimation compared with non-corrected images) were compared. In PETMRAC images, a volume of interest was drawn in the area of the artefact and in a reference site (contralateral body part); the mean and maximum standardised uptake values (SUVmean; SUVmax) were measured. RESULTS: Of 27 implanted materials (20 dental fillings, 3 injection ports, 3 hip prostheses and 1 sternal cerclage), 27 (100%) caused artefacts in CT, 19 (70%) in T1 weighted MRI and 17 (63%) in MRAC maps. 20 (74%) caused a visual overestimation of the (18)F-FDG uptake in PETCTAC, 2 (7%) caused an underestimation and 5 (19%) had no effect. In PETMRAC, 19 (70%) caused spherical extinctions and 8 (30%) had no effect. Mean values for SUVmean and SUVmax were significantly decreased in artefact-harbouring sites (p<0.001). CONCLUSION: Contrary to PET attenuation correction artefacts in PET/CT, which often show an overestimation of the (18)F-FDG uptake, MRAC artefacts owing to implanted medical materials in most cases cause an underestimation. ADVANCES IN KNOWLEDGE: Being aware of the morphology of artefacts owing to implanted medical materials avoids interpretation errors when reading PET/MRI.
| S-EPMC3635797 | biostudies-other