Mapping the association between tau-PET and Aβ-amyloid-PET using deep learning.
Ontology highlight
ABSTRACT:
SUBMITTER: Ruwanpathirana GP
PROVIDER: S-EPMC9427855 | biostudies-literature | 2022 Aug
REPOSITORIES: biostudies-literature
Publications
Mapping the association between tau-PET and Aβ-amyloid-PET using deep learning.
Ruwanpathirana Gihan P GP Williams Robert C RC Masters Colin L CL Rowe Christopher C CC Johnston Leigh A LA Davey Catherine E CE
Scientific reports 20220830 1
In Alzheimer's disease, the molecular pathogenesis of the extracellular Aβ-amyloid (Aβ) instigation of intracellular tau accumulation is poorly understood. We employed a high-resolution PET scanner, with low detection thresholds, to examine the Aβ-tau association using a convolutional neural network (CNN), and compared results to a standard voxel-wise linear analysis. The full range of Aβ Centiloid values was highly predicted by the tau topography using the CNN (training R<sup>2</sup> = 0.86, va ...[more]
Similar Datasets
Project description:PurposePET measures of amyloid and tau pathologies are powerful biomarkers for the diagnosis and monitoring of Alzheimer's disease (AD). Because cortical regions are close to bone, quantitation accuracy of amyloid and tau PET imaging can be significantly influenced by errors of attenuation correction (AC). This work presents an MR-based AC method that combines deep learning with a novel ultrashort time-to-echo (UTE)/multi-echo Dixon (mUTE) sequence for amyloid and tau imaging.MethodsThirty-five subjects that underwent both 11C-PiB and 18F-MK6240 scans were included in this study. The proposed method was compared with Dixon-based atlas method as well as magnetization-prepared rapid acquisition with gradient echo (MPRAGE)- or Dixon-based deep learning methods. The Dice coefficient and validation loss of the generated pseudo-CT images were used for comparison. PET error images regarding standardized uptake value ratio (SUVR) were quantified through regional and surface analysis to evaluate the final AC accuracy.ResultsThe Dice coefficients of the deep learning methods based on MPRAGE, Dixon, and mUTE images were 0.84 (0.91), 0.84 (0.92), and 0.87 (0.94) for the whole-brain (above-eye) bone regions, respectively, higher than the atlas method of 0.52 (0.64). The regional SUVR error for the atlas method was around 6%, higher than the regional SUV error. The regional SUV and SUVR errors for all deep learning methods were below 2%, with mUTE-based deep learning method performing the best. As for the surface analysis, the atlas method showed the largest error (> 10%) near vertices inside superior frontal, lateral occipital, superior parietal, and inferior temporal cortices. The mUTE-based deep learning method resulted in the least number of regions with error higher than 1%, with the largest error (> 5%) showing up near the inferior temporal and medial orbitofrontal cortices.ConclusionDeep learning with mUTE can generate accurate AC for amyloid and tau imaging in PET/MR.
| S-EPMC8411350 | biostudies-literature
Project description:Background and purposeThere is increasing recognition of the importance of cortical microinfarcts to overall brain health, cognition, and Alzheimer dementia. Cerebral small vessel pathologies are associated with microinfarcts and frequently coexist with Alzheimer disease; however, the extent to which Aβ (amyloid beta) and tau pathology modulates microvascular pathogenesis is not fully understood. Study objective was to examine the relationship of small vessel pathologies, arteriolosclerosis, and cerebral amyloid angiopathy, with cortical microinfarcts in people with differing levels of Aβ or tau tangle burden.MethodsParticipants were 1489 autopsied older people (mean age at death, 89 years; 67% women) from 1 of 3 ongoing clinical-pathological cohort studies of aging. Neuropathological evaluation identified cortical Aβ and tau tangle burden using immunohistochemistry in 8 brain regions, provided semiquantitative grading of cerebral vessel pathologies, and identified the presence of cortical microinfarcts. Logistic regression models adjusted for demographics and atherosclerosis and examined whether Aβ or tau tangle burden modified relations between small vessel pathologies and cortical microinfarcts.ResultsCortical microinfarcts were present in 17% of older people, moderate-to-severe cerebral amyloid angiopathy pathology in 36%, and arteriolosclerosis in 34%. In logistic regression models, we found interactions with Aβ and tau tangles, reflecting that the association between arteriolosclerosis and cortical microinfarcts was stronger in the context of greater Aβ (estimate, 0.15; SE=0.07; P=0.02) and tau tangle burden (estimate, 0.13; SE=0.06; P=0.02). Interactions also emerged for cerebral amyloid angiopathy, suggesting that the association between cerebral amyloid angiopathy and cortical microinfarcts is more robust in the presence of higher Aβ (estimate, 0.27; SE=0.07; P<0.001) and tangle burden (estimate, 0.16; SE=0.06; P=0.005).ConclusionsThese findings suggest that in the presence of elevated Aβ or tangle pathology, small vessel pathologies are associated with greater microvascular tissue injury, highlighting a potential link between neurodegenerative and vascular mechanisms.
| S-EPMC7902459 | biostudies-literature
Project description:PurposeWe aimed to evaluate the performance of deep learning-based generalization of ultra-low-count amyloid PET/MRI enhancement when applied to studies acquired with different scanning hardware and protocols.MethodsEighty simultaneous [18F]florbetaben PET/MRI studies were acquired, split equally between two sites (site 1: Signa PET/MRI, GE Healthcare, 39 participants, 67 ± 8 years, 23 females; site 2: mMR, Siemens Healthineers, 64 ± 11 years, 23 females) with different MRI protocols. Twenty minutes of list-mode PET data (90-110 min post-injection) were reconstructed as ground-truth. Ultra-low-count data obtained from undersampling by a factor of 100 (site 1) or the first minute of PET acquisition (site 2) were reconstructed for ultra-low-dose/ultra-short-time (1% dose and 5% time, respectively) PET images. A deep convolution neural network was pre-trained with site 1 data and either (A) directly applied or (B) trained further on site 2 data using transfer learning. Networks were also trained from scratch based on (C) site 2 data or (D) all data. Certified physicians determined amyloid uptake (+/-) status for accuracy and scored the image quality. The peak signal-to-noise ratio, structural similarity, and root-mean-squared error were calculated between images and their ground-truth counterparts. Mean regional standardized uptake value ratios (SUVR, reference region: cerebellar cortex) from 37 successful site 2 FreeSurfer segmentations were analyzed.ResultsAll network-synthesized images had reduced noise than their ultra-low-count reconstructions. Quantitatively, image metrics improved the most using method B, where SUVRs had the least variability from the ground-truth and the highest effect size to differentiate between positive and negative images. Method A images had lower accuracy and image quality than other methods; images synthesized from methods B-D scored similarly or better than the ground-truth images.ConclusionsDeep learning can successfully produce diagnostic amyloid PET images from short frame reconstructions. Data bias should be considered when applying pre-trained deep ultra-low-count amyloid PET/MRI networks for generalization.
| S-EPMC7680289 | biostudies-literature
Project description:ObjectiveTo determine if PET-based stages of regional amyloid deposition are associated with neuropathological phases of Aβ pathology.MethodsWe applied data-driven regional frequency-based and a-priori striatum-based PET staging approaches to ante-mortem 18F-Florbetapir-PET scans of 30 cases from the Alzheimer's Disease Neuroimaging Initiative autopsy cohort, and used Bayesian regression analysis to study the associations of these in vivo amyloid stages with neuropathological Thal phases of regional Aβ plaque distribution and with semi-quantitative ratings of neocortical and striatal plaque densities.ResultsBayesian regression revealed extreme evidence for an association of both PET-based staging approaches with Thal phases, and these associations were about 44 times more likely for frequency-based stages and 89 times more likely for striatum-based stages than for global cortical 18F-Florbetapir-PET signal. Early (i.e., neocortical-only) PET-based amyloid stages also predicted the absence of striatal/diencephalic cored plaques. Receiver operating characteristics curves revealed highly accurate discrimination between low/high Thal phases and the presence/absence of regional plaques. The median areas under the curve were 0.99 for frequency-based staging (95% credibility interval 0.97-1.00), 0.93 for striatum-based staging (0.83-1.00), and 0.87 for global 18F-Florbetapir-PET signal (0.72-0.98).InterpretationOur data indicate that both regional frequency- and striatum-based amyloid-PET staging approaches were superior to standard global amyloid-PET signal for differentiating between low and high degrees of regional amyloid pathology spread. Despite this, we found no evidence for the ability of either staging scheme to differentiate between low and moderate degrees of amyloid pathology which may be particularly relevant for early, preclinical stages of Alzheimer's disease.
| S-EPMC7818279 | biostudies-literature
Project description:PurposeImage quality of positron emission tomography (PET) is limited by various physical degradation factors. Our study aims to perform PET image denoising by utilizing prior information from the same patient. The proposed method is based on unsupervised deep learning, where no training pairs are needed.MethodsIn this method, the prior high-quality image from the patient was employed as the network input and the noisy PET image itself was treated as the training label. Constrained by the network structure and the prior image input, the network was trained to learn the intrinsic structure information from the noisy image and output a restored PET image. To validate the performance of the proposed method, a computer simulation study based on the BrainWeb phantom was first performed. A 68Ga-PRGD2 PET/CT dataset containing 10 patients and a 18F-FDG PET/MR dataset containing 30 patients were later on used for clinical data evaluation. The Gaussian, non-local mean (NLM) using CT/MR image as priors, BM4D, and Deep Decoder methods were included as reference methods. The contrast-to-noise ratio (CNR) improvements were used to rank different methods based on Wilcoxon signed-rank test.ResultsFor the simulation study, contrast recovery coefficient (CRC) vs. standard deviation (STD) curves showed that the proposed method achieved the best performance regarding the bias-variance tradeoff. For the clinical PET/CT dataset, the proposed method achieved the highest CNR improvement ratio (53.35% ± 21.78%), compared with the Gaussian (12.64% ± 6.15%, P = 0.002), NLM guided by CT (24.35% ± 16.30%, P = 0.002), BM4D (38.31% ± 20.26%, P = 0.002), and Deep Decoder (41.67% ± 22.28%, P = 0.002) methods. For the clinical PET/MR dataset, the CNR improvement ratio of the proposed method achieved 46.80% ± 25.23%, higher than the Gaussian (18.16% ± 10.02%, P < 0.0001), NLM guided by MR (25.36% ± 19.48%, P < 0.0001), BM4D (37.02% ± 21.38%, P < 0.0001), and Deep Decoder (30.03% ± 20.64%, P < 0.0001) methods. Restored images for all the datasets demonstrate that the proposed method can effectively smooth out the noise while recovering image details.ConclusionThe proposed unsupervised deep learning framework provides excellent image restoration effects, outperforming the Gaussian, NLM methods, BM4D, and Deep Decoder methods.
| S-EPMC7814987 | biostudies-literature
Project description:ObjectiveTo investigate the association between discordant β-amyloid (Aβ) PET and CSF biomarkers at baseline and the emergence of tau pathology 5 years later.MethodsWe included 730 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants without dementia (282 cognitively normal, 448 mild cognitive impairment) with baseline [18F]florbetapir PET and CSF Aβ42 available. Aβ CSF/PET status was determined at baseline using established cutoffs. Longitudinal data were available for [18F]florbetapir (Aβ) PET (baseline to 4.3 ± 1.9 years), CSF (p)tau (baseline to 2.0 ± 0.1 years), cognition (baseline to 4.3 ± 2.0 years), and [18F]flortaucipir (tau) PET (measured 5.2 ± 1.2 years after baseline to 1.6 ± 0.7 years later). We used linear mixed modeling to study the association between Aβ CSF/PET status and tau pathology measured in CSF or using PET. We calculated the proportion of CSF+/PET- participants who during follow-up (1) progressed to Aβ CSF+/PET+ or (2) became tau-positive based on [18F]flortaucipir PET.ResultsAβ CSF+/PET+ (n = 318) participants had elevated CSF (p)tau levels and worse cognitive performance at baseline, while CSF+/PET- (n = 80) participants were overall similar to the CSF-/PET- (N = 306) group. Five years after baseline, [18F]flortaucipir PET uptake in the CSF+/PET- group (1.20 ± 0.13) did not differ from CSF-/PET- (1.18 ± 0.08, p = 0.69), but was substantially lower than CSF+/PET+ (1.48 ± 0.44, p < 0.001). Of the CSF+/PET- participants, 21/64 (33%) progressed to Aβ CSF+/PET+, whereas only one (3%, difference p < 0.05) became tau-positive based on [18F]flortaucipir PET.ConclusionsAβ load detectable by both CSF and PET seems to precede substantial tau deposition. Compared to participants with abnormal Aβ levels on both PET and CSF, the CSF+/PET- group has a distinctly better prognosis.
| S-EPMC7963352 | biostudies-literature
Project description:We examined the characteristics of individuals with biomarker evidence of tauopathy but without β-amyloid (Aβ) (A-T+) in relation to individuals with (A+T+) and without (A-T-) evidence of Alzheimer's disease (AD). We included 561 participants with Aβ and tau PET from the Alzheimer's Disease Neuroimaging Initiative (ADNI). We compared A-T- (n = 316), A-T+ (n = 63), and A+T+ (n = 182) individuals on demographics, amyloid, tau, hippocampal volumes, and cognition. A-T+ individuals were low on apolipoprotein E ɛ4 prevalence (17%) and had no evidence of subtly elevated brain Aβ within the negative range. The severity of tau deposition, hippocampal atrophy, and cognitive dysfunction in the A-T+ group was intermediate between A-T- and A+T+ (all p < 0.001). Tau uptake patterns in A-T+ individuals were heterogeneous, but approximately 29% showed tau deposition in the medial temporal lobe only, consistent with primary age-related tauopathy and an additional 32% showed a pattern consistent with AD. A-T+ individuals also share other features that are characteristic of AD such as cognitive impairment and neurodegeneration, but this group is heterogeneous and likely reflects more than one disorder.
| S-EPMC9695003 | biostudies-literature
Project description:Staging the severity of Alzheimer's disease pathology using biomarkers is useful for therapeutic trials and clinical prognosis. Disease staging with amyloid and tau PET has face validity; however, this would be more practical with plasma biomarkers. Our objectives were, first, to examine approaches for staging amyloid and tau PET and, second, to examine prediction of amyloid and tau PET stages using plasma biomarkers. Participants (n = 1136) were enrolled in either the Mayo Clinic Study of Aging or the Alzheimer's Disease Research Center; had a concurrent amyloid PET, tau PET and blood draw; and met clinical criteria for cognitively unimpaired (n = 864), mild cognitive impairment (n = 148) or Alzheimer's clinical syndrome with dementia (n = 124). The latter two groups were combined into a cognitively impaired group (n = 272). We used multinomial regression models to estimate discrimination [concordance (C) statistics] among three amyloid PET stages (low, intermediate, high), four tau PET stages (Braak 0, 1-2, 3-4, 5-6) and a combined amyloid and tau PET stage (none/low versus intermediate/high severity) using plasma biomarkers as predictors separately within unimpaired and impaired individuals. Plasma analytes, p-tau181, Aβ1-42 and Aβ1-40 (analysed as the Aβ42/Aβ40 ratio), glial fibrillary acidic protein and neurofilament light chain were measured on the HD-X Simoa Quanterix platform. Plasma p-tau217 was also measured in a subset (n = 355) of cognitively unimpaired participants using the Lilly Meso Scale Discovery assay. Models with all Quanterix plasma analytes along with risk factors (age, sex and APOE) most often provided the best discrimination among amyloid PET stages (C = 0.78-0.82). Models with p-tau181 provided similar discrimination of tau PET stages to models with all four plasma analytes (C = 0.72-0.85 versus C = 0.73-0.86). Discriminating a PET proxy of intermediate/high from none/low Alzheimer's disease neuropathological change with all four Quanterix plasma analytes was excellent but not better than p-tau181 only (C = 0.88 versus 0.87 for unimpaired and C = 0.91 versus 0.90 for impaired). Lilly p-tau217 outperformed the Quanterix p-tau181 assay for discriminating high versus intermediate amyloid (C = 0.85 versus 0.74) but did not improve over a model with all Quanterix plasma analytes and risk factors (C = 0.85 versus 0.83). Plasma analytes along with risk factors can discriminate between amyloid and tau PET stages and between a PET surrogate for intermediate/high versus none/low neuropathological change with accuracy in the acceptable to excellent range. Combinations of plasma analytes are better than single analytes for many staging predictions with the exception that Quanterix p-tau181 alone usually performed equivalently to combinations of Quanterix analytes for tau PET discrimination.
| S-EPMC10151195 | biostudies-literature
Project description:IntroductionPositron emission tomography (PET) imaging targeting neurofibrillary tau tangles is increasingly used in the study of Alzheimer's disease (AD), but its utility may be limited by conventional quantitative or qualitative evaluation techniques in earlier disease states. Convolutional neural networks (CNNs) are effective in learning spatial patterns for image classification.Methods18F-MK6240 (n = 320) and AV-1451 (n = 446) PET images were pooled from multiple studies. We performed iterations with differing permutations of radioligands, heuristics, and architectures. Performance was compared to a standard region of interest (ROI)-based approach on prediction of memory impairment. We visualized attention of the network to illustrate decision making.ResultsOverall, models had high accuracy (> 80%) with good average sensitivity and specificity (75% and 82%, respectively), and had comparable or higher accuracy to the ROI standard. Visualizations of model attention highlight known characteristics of tau radioligand binding.DiscussionCNNs could improve tau PET's role in early disease and extend the utility of tau PET across generations of radioligands.
| S-EPMC8719427 | biostudies-literature
Project description:IntroductionAlzheimer's disease (AD) pathology causes corticobasal syndrome (CBS) in 21%-50% of patients. Studies have assessed hypometabolism in CBS according to β-amyloid (A) positron emission tomography (PET), but the understanding of the association of both AD-tau (T) and A with hypometabolism is incomplete.MethodsThirty-three CBS patients and 45 controls underwent fluorodeoxyglucose (FDG), flortaucipir, and Pittsburgh compound-B PET and were classified as A± and T±. FDG-PET uptake was extracted for 12 regions-of-interest in dominant (most affected) and non-dominant hemispheres and compared across A/T groups.ResultsA+T+ patients had greater hypometabolism in temporo-parieto-occipital cortices than A+T- and A-T- groups, with no differences observed between the A+T- and A-T- groups. FDG asymmetry was more accentuated in A+T+ patients. Medial temporal and basal ganglia metabolism were similar across AT groups.DiscussionAmyloid and tau positivity contribute synergistically to hypometabolism and asymmetry in temporo-parieto-occipital cortices in CBS, with AD-like patterns of hypometabolism observed only in A+T+ patients.HighlightsAmyloid (A) and tau PET (T) status can be used to stratify CBS patients. A+T+ CBS patients show more hypometabolism in temporo-parieto-occipital cortices. Medial temporal metabolism (typical AD pattern) is similar across AT groups. Parieto-occipital cortices should be assessed when investigating AT pathology in CBS. Amyloid and tau positivity contribute synergistically to hypometabolism and asymmetry in CBS.
| S-EPMC11923383 | biostudies-literature