On the accuracy and reproducibility of a novel probabilistic atlas-based generation for calculation of head attenuation maps on integrated PET/MR scanners

Accuracy and Longitudinal Consistency of PET/MR Attenuation Correction in Amyloid PET Imaging amid Software and Hardware Upgrades

BACKGROUND AND PURPOSE

Integrated PET/MR allows the simultaneous acquisition of PET biomarkers and structural and functional MRI to study Alzheimer disease (AD). Attenuation correction (AC), crucial for PET quantification, can be performed by using a deep learning approach, DL-Dixon, based on standard Dixon images. Longitudinal amyloid PET imaging, which provides important information about disease progression or treatment responses in AD, is usually acquired over several years. Hardware and software upgrades often occur during a multiple-year study period, resulting in data variability. This study aims to harmonize PET/MR DL-Dixon AC amid software and head coil updates and evaluate its accuracy and longitudinal consistency.

MATERIALS AND METHODS

Tri-modality PET/MR and CT images were obtained from 329 participants, with a subset of 38 undergoing tri-modality scans twice within approximately 3 years. Transfer learning was used to fine-tune DL-Dixon models on images from 2 scanner software versions (VB20P and VE11P) and 2 head coils (16-channel and 32-channel coils). The accuracy and longitudinal consistency of the DL-Dixon AC were evaluated. Power analyses were performed to estimate the sample size needed to detect various levels of longitudinal changes in the PET standardized uptake value ratio (SUVR).

RESULTS

The DL-Dixon method demonstrated high accuracy across all data, irrespective of scanner software versions and head coils. More than 95.6% of brain voxels showed less than 10% PET relative absolute error in all participants. The median [interquartile range] PET mean relative absolute error was 1.10% [0.93%, 1.26%], 1.24% [1.03%, 1.54%], 0.99% [0.86%, 1.13%] in the cortical summary region, and 1.04% [0.83%, 1.36%], 1.08% [0.84%, 1.34%], 1.05% [0.72%, 1.32%] in cerebellum by using the DL-Dixon models for the VB20P 16-channel coil, VE11P 16-channel coil, and VE11P 32-channel coil data, respectively. The within-subject coefficient of variation and intraclass correlation coefficient of PET SUVR in the cortical regions were comparable between the DL-Dixon and CT AC. Power analysis indicated that similar numbers of participants would be needed to detect the same level of PET changes by using DL-Dixon and CT AC.

CONCLUSIONS

DL-Dixon exhibited excellent accuracy and longitudinal consistency across the 2 software versions and head coils, demonstrating its robustness for longitudinal PET/MR neuroimaging studies in AD.

Update on the Use of PET/MRI Contrast Agents and Tracers in Brain Oncology: A Systematic Review

The recent advancements in hybrid positron emission tomography–magnetic resonance imaging systems (PET/MRI) have brought massive value in the investigation of disease processes, in the development of novel treatments, in the monitoring of both therapy response and disease progression, and, not least, in the introduction of new multidisciplinary molecular imaging approaches. While offering potential advantages over PET/CT, the hybrid PET/MRI proved to improve both the image quality and lesion detectability. In particular, it showed to be an effective tool for the study of metabolic information about lesions and pathological conditions affecting the brain, from a better tumor characterization to the analysis of metabolic brain networks. Based on the PRISMA guidelines, this work presents a systematic review on PET/MRI in basic research and clinical differential diagnosis on brain oncology and neurodegenerative disorders. The analysis includes literature works and clinical case studies, with a specific focus on the use of PET tracers and MRI contrast agents, which are usually employed to perform hybrid PET/MRI studies of brain tumors. A systematic literature search for original diagnostic studies is performed using PubMed/MEDLINE, Scopus and Web of Science. Patients, study, and imaging characteristics were extracted from the selected articles. The analysis included acquired data pooling, heterogeneity testing, sensitivity analyses, used tracers, and reported patient outcomes. Our analysis shows that, while PET/MRI for the brain is a promising diagnostic method for early diagnosis, staging and recurrence in patients with brain diseases, a better definition of the role of tracers and imaging agents in both clinical and preclinical hybrid PET/MRI applications is needed and further efforts should be devoted to the standardization of the contrast imaging protocols, also considering the emerging agents and multimodal probes.

Evaluation of three methods for delineation and attenuation estimation of the sinus region in MR-based attenuation correction for brain PET-MR imaging

Background

Attenuation correction is crucial in quantitative positron emission tomography-magnetic resonance (PET-MRI) imaging. We evaluated three methods to improve the segmentation and modelling of the attenuation coefficients in the nasal sinus region. Two methods (cuboid and template method) included a MRI-CT conversion model for assigning the attenuation coefficients in the nasal sinus region, whereas one used fixed attenuation coefficient assignment (bulk method).

Methods

The study population consisted of data of 10 subjects which had undergone PET-CT and PET-MRI. PET images were reconstructed with and without time-of-flight (TOF) using CT-based attenuation correction (CTAC) as reference. Comparison was done visually, using DICE coefficients, correlation, analyzing attenuation coefficients, and quantitative analysis of PET and bias atlas images.

Results

The median DICE coefficients were 0.824, 0.853, 0.849 for the bulk, cuboid and template method, respectively. The median attenuation coefficients were 0.0841 cm⁻¹, 0.0876 cm⁻¹, 0.0861 cm⁻¹ and 0.0852 cm⁻¹, for CTAC, bulk, cuboid and template method, respectively. The cuboid and template methods showed error of less than 2.5% in attenuation coefficients. An increased correlation to CTAC was shown with the cuboid and template methods. In the regional analysis, improvement in at least 49% and 80% of VOI was seen with non-TOF and TOF imaging. All methods showed errors less than 2.5% in non-TOF and less than 2% in TOF reconstructions.

Conclusions

We evaluated two proof-of-concept methods for improving quantitative accuracy in PET/MRI imaging and showed that bias can be further reduced by inclusion of TOF. Largest improvements were seen in the regions of olfactory bulb, Heschl's gyri, lingual gyrus and cerebellar vermis. However, the overall effect of inclusion of the sinus region as separate class in MRAC to PET quantification in the brain was considered modest.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12880-022-00770-0.

Evaluating different methods of MR-based motion correction in simultaneous PET/MR using a head phantom moved by a robotic system

Background

Due to comparatively long measurement times in simultaneous positron emission tomography and magnetic resonance (PET/MR) imaging, patient movement during the measurement can be challenging. This leads to artifacts which have a negative impact on the visual assessment and quantitative validity of the image data and, in the worst case, can lead to misinterpretations. Simultaneous PET/MR systems allow the MR-based registration of movements and enable correction of the PET data. To assess the effectiveness of motion correction methods, it is necessary to carry out measurements on phantoms that are moved in a reproducible way. This study explores the possibility of using such a phantom-based setup to evaluate motion correction strategies in PET/MR of the human head.

Method

An MR-compatible robotic system was used to generate rigid movements of a head-like phantom. Different tools, either from the manufacturer or open-source software, were used to estimate and correct for motion based on the PET data itself (SIRF with SPM and NiftyReg) and MR data acquired simultaneously (e.g. MCLFIRT, BrainCompass). Different motion estimates were compared using data acquired during robot-induced motion. The effectiveness of motion correction of PET data was evaluated by determining the segmented volume of an activity-filled flask inside the phantom. In addition, the segmented volume was used to determine the centre-of-mass and the change in maximum activity concentration.

Results

The results showed a volume increase between 2.7 and 36.3% could be induced by the experimental setup depending on the motion pattern. Both, BrainCompass and MCFLIRT, produced corrected PET images, by reducing the volume increase to 0.7–4.7% (BrainCompass) and to -2.8–0.4% (MCFLIRT). The same was observed for example for the centre-of-mass, where the results show that MCFLIRT (0.2–0.6 mm after motion correction) had a smaller deviation from the reference position than BrainCompass (0.5–1.8 mm) for all displacements.

Conclusions

The experimental setup is suitable for the reproducible generation of movement patterns. Using open-source software for motion correction is a viable alternative to the vendor-provided motion-correction software.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40658-022-00442-6.

Deep learning-based T1-enhanced selection of linear attenuation coefficients (DL-TESLA) for PET/MR attenuation correction in dementia neuroimaging

PURPOSE

The accuracy of existing PET/MR attenuation correction (AC) has been limited by a lack of correlation between MR signal and tissue electron density. Based on our finding that longitudinal relaxation rate, or R1 , is associated with CT Hounsfield unit in bone and soft tissues in the brain, we propose a deep learning T1 -enhanced selection of linear attenuation coefficients (DL-TESLA) method to incorporate quantitative R1 for PET/MR AC and evaluate its accuracy and longitudinal test-retest repeatability in brain PET/MR imaging.

METHODS

DL-TESLA uses a 3D residual UNet (ResUNet) for pseudo-CT (pCT) estimation. With a total of 174 participants, we compared PET AC accuracy of DL-TESLA to 3 other methods adopting similar 3D ResUNet structures but using UTE R 2 ∗ , or Dixon, or T1 -MPRAGE as input. With images from 23 additional participants repeatedly scanned, the test-retest differences and within-subject coefficient of variation of standardized uptake value ratios (SUVR) were compared between PET images reconstructed using either DL-TESLA or CT for AC.

RESULTS

DL-TESLA had (1) significantly lower mean absolute error in pCT, (2) the highest Dice coefficients in both bone and air, (3) significantly lower PET relative absolute error in whole brain and various brain regions, (4) the highest percentage of voxels with a PET relative error within both ±3% and ±5%, (5) similar to CT test-retest differences in SUVRs from the cerebrum and mean cortical (MC) region, and (6) similar to CT within-subject coefficient of variation in cerebrum and MC.

CONCLUSION

DL-TESLA demonstrates excellent PET/MR AC accuracy and test-retest repeatability.

Attenuation correction for human PET/MRI studies

Attenuation correction has been one of the main methodological challenges in the integrated positron emission tomography and magnetic resonance imaging (PET/MRI) field. As standard transmission or computed tomography approaches are not available in integrated PET/MRI scanners, MR-based attenuation correction approaches had to be developed. Aspects that have to be considered for implementing accurate methods include the need to account for attenuation in bone tissue, normal and pathological lung and the MR hardware present in the PET field-of-view, to reduce the impact of subject motion, to minimize truncation and susceptibility artifacts, and to address issues related to the data acquisition and processing both on the PET and MRI sides. The standard MR-based attenuation correction techniques implemented by the PET/MRI equipment manufacturers and their impact on clinical and research PET data interpretation and quantification are first discussed. Next, the more advanced methods, including the latest generation deep learning-based approaches that have been proposed for further minimizing the attenuation correction related bias are described. Finally, a future perspective focused on the needed developments in the field is given.

MR Image-Based Attenuation Correction of Brain PET Imaging: Review of Literature on Machine Learning Approaches for Segmentation

Recent emerging hybrid technology of positron emission tomography/magnetic resonance (PET/MR) imaging has generated a great need for an accurate MR image-based PET attenuation correction. MR image segmentation, as a robust and simple method for PET attenuation correction, has been clinically adopted in commercial PET/MR scanners. The general approach in this method is to segment the MR image into different tissue types, each assigned an attenuation constant as in an X-ray CT image. Machine learning techniques such as clustering, classification and deep networks are extensively used for brain MR image segmentation. However, only limited work has been reported on using deep learning in brain PET attenuation correction. In addition, there is a lack of clinical evaluation of machine learning methods in this application. The aim of this review is to study the use of machine learning methods for MR image segmentation and its application in attenuation correction for PET brain imaging. Furthermore, challenges and future opportunities in MR image-based PET attenuation correction are discussed.

Full-count PET recovery from low-count image using a dilated convolutional neural network

PURPOSE

Positron emission tomography (PET) is an essential technique in many clinical applications that allows for quantitative imaging at the molecular level. This study aims to develop a denoising method using a novel dilated convolutional neural network (CNN) to recover full-count images from low-count images.

METHODS

We adopted similar hierarchical structures as the conventional U-Net and incorporated dilated kernels in each convolution to allow the network to observe larger, more robust features within the image without the requirement of downsampling and upsampling internal representations. Our dNet was trained alongside a U-Net for comparison. Both models were evaluated using a leave-one-out cross-validation procedure on a dataset of 35 subjects (~3500 slabs), which were obtained from an ongoing ¹⁸ F-Fluorodeoxyglucose (FDG) study. Low-count PET data (10% count) were generated by randomly selecting one-tenth of all events in the associated listmode file. Analysis was done on the static image from the last 10 minutes of emission data. Both low-count PET and full-count PET were reconstructed using ordered subset expectation maximization (OSEM). Objective image quality metrics, including mean absolute percent error (MAPE), peak signal-to-noise ratio (PSNR), and structural similarity index metric (SSIM), were used to analyze the deep learning methods. Both deep learning methods were further compared to a traditional Gaussian filtering method. Further, region of interest (ROI) quantitative analysis was also used to compare U-Net and dNet architectures.

RESULTS

Both the U-Net and our proposed network were successfully trained to synthesize full-count PET images from the generated low-count PET images. Compared to low-count PET and Gaussian filtering, both deep learning methods improved MAPE, PSNR, and SSIM. Our dNet also systematically outperformed U-Net on all three metrics (MAPE: 4.99 ± 0.68 vs 5.31 ± 0.76, P < 0.01; PSNR: 31.55 ± 1.31 dB vs 31.05 ± 1.39, P < 0.01; SSIM: 0.9513 ± 0.0154 vs 0.9447 ± 0.0178, P < 0.01). ROI quantification showed greater quantitative improvements using dNet over U-Net.

CONCLUSION

This study proposed a novel approach of using dilated convolutions for recovering full-count PET images from low-count PET images.

Hybrid PET- and MR-driven attenuation correction for enhanced 18F-NaF and 18F-FDG quantification in cardiovascular PET/MR imaging

BACKGROUND

The standard MR Dixon-based attenuation correction (AC) method in positron emission tomography/magnetic resonance (PET/MR) imaging segments only the air, lung, fat and soft-tissues (4-class), thus neglecting the highly attenuating bone tissues and affecting quantification in bones and adjacent vessels. We sought to address this limitation by utilizing the distinctively high bone uptake rate constant Ki expected from 18F-Sodium Fluoride (18F-NaF) to segment bones from PET data and support 5-class hybrid PET/MR-driven AC for 18F-NaF and 18F-Fluorodeoxyglucose (18F-FDG) PET/MR cardiovascular imaging.

METHODS

We introduce 5-class Ki/MR-AC for (i) 18F-NaF studies where the bones are segmented from Patlak Ki images and added as the 5th tissue class to the MR Dixon 4-class AC map. Furthermore, we propose two alternative dual-tracer protocols to permit 5-class Ki/MR-AC for (ii) 18F-FDG-only data, with a streamlined simultaneous administration of 18F-FDG and 18F-NaF at 4:1 ratio (R4:1), or (iii) for 18F-FDG-only or both 18F-FDG and 18F-NaF dual-tracer data, by administering 18F-NaF 90 minutes after an equal 18F-FDG dosage (R1:1). The Ki-driven bone segmentation was validated against computed tomography (CT)-based segmentation in rabbits, followed by PET/MR validation on 108 vertebral bone and carotid wall regions in 16 human volunteers with and without prior indication of carotid atherosclerosis disease (CAD).

RESULTS

In rabbits, we observed similar (< 1.2% mean difference) vertebral bone 18F-NaF SUVmean scores when applying 5-class AC with Ki-segmented bone (5-class Ki/CT-AC) vs CT-segmented bone (5-class CT-AC) tissue. Considering the PET data corrected with continuous CT-AC maps as gold-standard, the percentage SUVmean bias was reduced by 17.6% (18F-NaF) and 15.4% (R4:1) with 5-class Ki/CT-AC vs 4-class CT-AC. In humans without prior CAD indication, we reported 17.7% and 20% higher 18F-NaF target-to-background ratio (TBR) at carotid bifurcations wall and vertebral bones, respectively, with 5- vs 4-class AC. In the R4:1 human cohort, the mean 18F-FDG:18F-NaF TBR increased by 12.2% at carotid bifurcations wall and 19.9% at vertebral bones. For the R1:1 cohort of subjects without CAD indication, mean TBR increased by 15.3% (18F-FDG) and 15.5% (18F-NaF) at carotid bifurcations and 21.6% (18F-FDG) and 22.5% (18F-NaF) at vertebral bones. Similar TBR enhancements were observed when applying the proposed AC method to human subjects with prior CAD indication.

CONCLUSIONS

Ki-driven bone segmentation and 5-class hybrid PET/MR-driven AC is feasible and can significantly enhance 18F-NaF and 18F-FDG contrast and quantification in bone tissues and carotid walls.

Proton range shift analysis on brain pseudo-CT generated from T1 and T2 MR

BACKGROUND

In radiotherapy, MR imaging is only used because it has significantly better soft tissue contrast than CT, but it lacks electron density information needed for dose calculation. This work assesses the feasibility of using pseudo-CT (pCT) generated from T1w/T2w MR for proton treatment planning, where proton range comparisons are performed between standard CT and pCT.

MATERIAL AND METHODS

MR and CT data from 14 glioblastoma patients were used in this study. The pCT was generated by using conversion libraries obtained from tissue segmentation and anatomical regioning of the T1w/T2w MR. For each patient, a plan consisting of three 18 Gy beams was designed on the pCT, for a total of 42 analyzed beams. The plan was then transferred onto the CT that represented the ground truth. Range shift (RS) between pCT and CT was computed at R₈₀ over 10 slices. The acceptance threshold for RS was according to clinical guidelines of two institutions. A γ-index test was also performed on the total dose for each patient.

RESULTS

Mean absolute error and bias for the pCT were 124 ± 10 and -16 ± 26 Hounsfield Units (HU), respectively. The median and interquartile range of RS was 0.5 and 1.4 mm, with highest absolute value being 4.4 mm. Of the 42 beams, 40 showed RS less than the clinical range margin. The two beams with larger RS were both in the cranio-caudal direction and had segmentation errors due to the partial volume effect, leading to misassignment of the HU.

CONCLUSIONS

This study showed the feasibility of using T1w and T2w MRI to generate a pCT for proton therapy treatment, thus avoiding the use of a planning CT and allowing better target definition and possibilities for online adaptive therapies. Further improvements of the methodology are still required to improve the conversion from MRI intensities to HUs.

An Efficient Approach to Perform MR-assisted PET Data Optimization in Simultaneous PET/MR Neuroimaging Studies

A main advantage of PET is that it provides quantitative measures of the radiotracer concentration, but its accuracy is confounded by several factors, including attenuation, subject motion, and limited spatial resolution. Using the information from one simultaneously acquired morphological MR sequence with embedded navigators, we propose an efficient method called MR-assisted PET data optimization (MaPET) to perform attenuation correction (AC), motion correction, and anatomy-aided reconstruction. Methods: For attenuation correction, voxel-wise linear attenuation coefficient maps were generated using an SPM8-based approach method on the MR volume. The embedded navigators were used to derive head motion estimates for event-based PET motion correction. The anatomy provided by the MR volume was incorporated into the PET image reconstruction using a kernel-based method. Region-based analyses were carried out to assess the quality of images generated through various stages of PET data optimization. Results: The optimized PET images reconstructed with MaPET was superior in image quality compared to images reconstructed using only attenuation correction, with high SNR and low coefficient of variation (5.08 and 0.229 in a composite cortical region compared to 3.12 and 0.570). The optimized images were also shown using the Cohen's d metric to achieve a greater effect size in distinguishing cortical regions with hypometabolism from regions of preserved metabolism in each individual for different diagnosis groups. Conclusion: We have shown the spatiotemporally correlated data acquired using a single MR sequence can be used for PET attenuation, motion and partial volume effects corrections and the MaPET method may enable more accurate assessment of pathological changes in dementia and other brain disorders.

MR-assisted PET motion correction in simultaneous PET/MRI studies of dementia subjects

BACKGROUND

Subject motion in positron emission tomography (PET) studies leads to image blurring and artifacts; simultaneously acquired magnetic resonance imaging (MRI) data provides a means for motion correction (MC) in integrated PET/MRI scanners.

PURPOSE

To assess the effect of realistic head motion and MR-based MC on static [¹⁸ F]-fluorodeoxyglucose (FDG) PET images in dementia patients.

STUDY TYPE

Observational study.

POPULATION

Thirty dementia subjects were recruited.

FIELD STRENGTH/SEQUENCE

3T hybrid PET/MR scanner where EPI-based and T₁ -weighted sequences were acquired simultaneously with the PET data.

ASSESSMENT

Head motion parameters estimated from high temporal resolution MR volumes were used for PET MC. The MR-based MC method was compared to PET frame-based MC methods in which motion parameters were estimated by coregistering 5-minute frames before and after accounting for the attenuation-emission mismatch. The relative changes in standardized uptake value ratios (SUVRs) between the PET volumes processed with the various MC methods, without MC, and the PET volumes with simulated motion were compared in relevant brain regions.

STATISTICAL TESTS

The absolute value of the regional SUVR relative change was assessed with pairwise paired t-tests testing at the P = 0.05 level, comparing the values obtained through different MR-based MC processing methods as well as across different motion groups. The intraregion voxelwise variability of regional SUVRs obtained through different MR-based MC processing methods was also assessed with pairwise paired t-tests testing at the P = 0.05 level.

RESULTS

MC had a greater impact on PET data quantification in subjects with larger amplitude motion (higher than 18% in the medial orbitofrontal cortex) and greater changes were generally observed for the MR-based MC method compared to the frame-based methods. Furthermore, a mean relative change of ∼4% was observed after MC even at the group level, suggesting the importance of routinely applying this correction. The intraregion voxelwise variability of regional SUVRs was also decreased using MR-based MC. All comparisons were significant at the P = 0.05 level.

DATA CONCLUSION

Incorporating temporally correlated MR data to account for intraframe motion has a positive impact on the FDG PET image quality and data quantification in dementia patients.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2018;47:1288-1296.

Visual detection of regional brain hypometabolism in cognitively impaired patients is independent of positron emission tomography-magnetic resonance attenuation correction method

Fluorodeoxyglucose (FDG) positron emission tomography-magnetic resonance (PET/MR) is useful for the evaluation of cognitively-impaired patients. This study aims to assess two different attenuation correction (AC) methods (Dixon-MR and atlas-based) versus index-standard computed tomography (CT) AC for the visual interpretation of regional hypometabolism in patients with cognitive impairment. Two board-certified nuclear medicine physicians blindly scored brain region FDG hypometabolism as normal versus hypometabolic using two-dimensional (2D) and 3D FDG PET/MR images generated by MIM software. Regions were quantitatively assessed as normal versus mildly, moderately, or severely hypometabolic. Hypometabolism scores obtained using the different methods of AC were compared, and interreader, as well as intra-reader agreement, was assessed. Regional hypometabolism versus normal metabolism was correctly classified in 16 patients on atlas-based and Dixon-based AC map PET reconstructions (vs. CT reference AC) for 94% (90%–96% confidence interval [CI]) and 93% (89%–96% CI) of scored regions, respectively. The averaged sensitivity/specificity for detection of any regional hypometabolism was 95%/94% (P = 0.669) and 90%/91% (P = 0.937) for atlas-based and Dixon-based AC maps. Interreader agreement for detection of regional hypometabolism was high, with similar outcome assessments when using atlas- and Dixon-corrected PET data in 93% (Κ =0.82) and 93% (Κ =0.84) of regions, respectively. Intrareader agreement for detection of regional hypometabolism was high, with concordant outcome assessments when using atlas- and Dixon-corrected data in 93%/92% (Κ =0.79) and 92/93% (Κ =0.78). Despite the quantitative advantages of atlas-based AC in brain PET/MR, routine clinical Dixon AC yields comparable visual ratings of regional hypometabolism in the evaluation of cognitively impaired patients undergoing brain PET/MR and is similar in performance to CT-based AC. Therefore, Dixon AC is acceptable for the routine clinical evaluation of dementia syndromes.