SPICER: Self-supervised learning for MRI with automatic coil sensitivity estimation and reconstruction.

PURPOSE: To introduce a novel deep model-based architecture (DMBA), SPICER, that uses pairs of noisy and undersampled k-space measurements of the same object to jointly train a model for MRI reconstruction and automatic coil sensitivity estimation. METHODS: SPICER consists of two modules to simultaneously reconstructs accurate MR images and estimates high-quality coil sensitivity maps (CSMs). The first module, CSM estimation module, uses a convolutional neural network (CNN) to estimate CSMs from the raw measurements. The second module, DMBA-based MRI reconstruction module, forms reconstructed images from the input measurements and the estimated CSMs using both the physical measurement model and learned CNN prior. With the benefit of our self-supervised learning strategy, SPICER can be efficiently trained without any fully sampled reference data. RESULTS: We validate SPICER on both open-access datasets and experimentally collected data, showing that it can achieve state-of-the-art performance in highly accelerated data acquisition settings (up to 10 × $$ 10\times $$ ). Our results also highlight the importance of different modules of SPICER-including the DMBA, the CSM estimation, and the SPICER training loss-on the final performance of the method. Moreover, SPICER can estimate better CSMs than pre-estimation methods especially when the ACS data is limited. CONCLUSION: Despite being trained on noisy undersampled data, SPICER can reconstruct high-quality images and CSMs in highly undersampled settings, which outperforms other self-supervised learning methods and matches the performance of the well-known E2E-VarNet trained on fully sampled ground-truth data.

Vascular space occupancy asymmetric spin echo (VASO-ASE) for non-invasive quantification of cerebral oxygen extraction fraction.

PURPOSE: Asymmetric spin echo (ASE) MRI is a method for measuring regional oxygen extraction fraction (OEF); however, extravascular tissue models have been shown to under-estimate OEF. The hypothesis investigated here is that the addition of a vascular-space-occupancy (VASO) pre-pulse will more fully suppress blood water signal and provide global OEF values more consistent with physiological expectation and 15 O positron emission tomography (PET)-validated T2 -relaxation-under-spin-tagging (TRUST) OEF measures. METHODS: Healthy adults (n = 14; age = 27.7 ± 5.2 y; sex = 7/7 male/female) were scanned at 3.0T. Multi-echo ASE without inter-readout refocusing (ASERF- ), multi-echo ASE with inter-readout refocusing (ASERF+ ), and single-echo VASO-ASE were acquired twice each with common spatial resolution = 3.44 × 3.44 × 3.0 mm and τ = 0-20 ms (interval = 0.5 ms). TRUST was acquired twice sequentially for independent global OEF assessment (τCPMG  = 10 ms; effective TEs = 0, 40, 80, and 160 ms; spatial resolution = 3.4 × 3.4 × 5 mm). OEF intraclass-correlation-coefficients (ICC), summary statistics, and group-wise differences were assessed (Wilcoxon rank-sum; significance: two-sided p < 0.05). RESULTS: ASERF+ (OEF = 36.8 ± 1.9%) and VASO-ASE (OEF = 34.4 ± 2.3%) produced OEF values similar to TRUST (OEF = 36.5 ± 4.6%, human calibration model; OEF = 32.7 ± 4.9%, bovine calibration model); however, ASERF- yielded lower OEF (OEF = 26.1 ± 1.0%; p < 0.01) relative to TRUST. VASO-ASE (ICC = 0.61) yielded lower ICC compared to other ASE variants (ICC >0.89). CONCLUSION: VASO-ASE and TRUST provide similar OEF values; however, VASO-ASE spatial coverage and repeatability improvements are required.

Oxygen Metabolic Stress and White Matter Injury in Patients With Cerebral Small Vessel Disease.

BACKGROUND: Chronic hypoxia-ischemia is a putative mechanism underlying the development of white matter hyperintensities (WMH) and microstructural disruption in cerebral small vessel disease. WMH fall primarily within deep white matter (WM) watershed regions. We hypothesized that elevated oxygen extraction fraction (OEF), a signature of hypoxia-ischemia, would be detected in the watershed where WMH density is highest. We further hypothesized that OEF would be elevated in regions immediately surrounding WMH, at the leading edge of growth. METHODS: In this cross-sectional study conducted from 2016 to 2019 at an academic medical center in St Louis, MO, participants (age >50) with a range of cerebrovascular risk factors underwent brain magnetic resonance imaging using pseudocontinuous arterial spin labeling, asymmetric spin echo, fluid-attenuated inversion recovery and diffusion tensor imaging to measure cerebral blood flow (CBF), OEF, WMH, and WM integrity, respectively. We defined the physiologic watershed as a region where CBF was below the 10th percentile of mean WM CBF in a young healthy cohort. We conducted linear regression to evaluate the relationship between CBF and OEF with structural and microstructural WM injury defined by fluid-attenuated inversion recovery WMH and diffusion tensor imaging, respectively. We conducted ANOVA to determine if OEF was increased in proximity to WMH lesions. RESULTS: In a cohort of 42 participants (age 50-80), the physiologic watershed region spatially overlapped with regions of highest WMH lesion density. As CBF decreased and OEF increased, WMH density increased. Elevated watershed OEF was associated with greater WMH burden and microstructural disruption, after adjusting for vascular risk factors. In contrast, WM and watershed CBF were not associated with WMH burden or microstructural disruption. Moreover, OEF progressively increased while CBF decreased, in concentric contours approaching WMH lesions. CONCLUSIONS: Chronic hypoxia-ischemia in the watershed region may contribute to cerebral small vessel disease pathogenesis and development of WMH. Watershed OEF may hold promise as an imaging biomarker to identify individuals at risk for cerebral small vessel disease progression.

Silent Infarcts, White Matter Integrity, and Oxygen Metabolic Stress in Young Adults With and Without Sickle Cell Trait.

BACKGROUND: Individuals with sickle cell anemia have heightened risk of stroke and cognitive dysfunction. Given its high prevalence globally, whether sickle cell trait (SCT) is a risk factor for neurological injury has been of interest; however, data have been limited. We hypothesized that young, healthy adults with SCT would show normal cerebrovascular structure and hemodynamic function. METHODS: As a case-control study, young adults with (N=25, cases) and without SCT (N=24, controls) underwent brain magnetic resonance imaging to quantify brain volume, microstructural integrity (fractional anisotropy), silent cerebral infarcts (SCI), intracranial stenosis, and aneurysms. Pseudocontinuous arterial spin labeling and asymmetric spin echo sequences measured cerebral blood flow and oxygen extraction fraction, respectively, from which cerebral metabolic oxygen demand was calculated. Imaging metrics were compared between SCT cases and controls. SCI volume was correlated with baseline characteristics. RESULTS: Compared with controls, adults with SCT demonstrated similar normalized brain volumes (SCT 0.80 versus control 0.81, P=0.41), white matter fractional anisotropy (SCT 0.41 versus control 0.43, P=0.37), cerebral blood flow (SCT 62.04 versus control, 61.16 mL/min/100 g, P=0.67), oxygen extraction fraction (SCT 0.27 versus control 0.27, P=0.31), and cerebral metabolic oxygen demand (SCT 2.71 versus control 2.70 mL/min/100 g, P=0.96). One per cohort had an intracranial aneurysm. None had intracranial stenosis. The SCT cases and controls showed similar prevalence and volume of SCIs; however, in the subset of participants with SCIs, the SCT cases had greater SCI volume versus controls (0.29 versus 0.07 mL, P=0.008). Of baseline characteristics, creatinine was mildly elevated in the SCT cohort (0.9 versus 0.8 mg/dL, P=0.053) and correlated with SCI volume (ρ=0.49, P=0.032). In the SCT cohort, SCI distribution was similar to that of young adults with sickle cell anemia. CONCLUSIONS: Adults with SCT showed normal cerebrovascular structure and hemodynamic function. These findings suggest that healthy individuals with SCT are unlikely to be at increased risk for early or accelerated ischemic brain injury.

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 R2∗ , 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.

Effects of a brief HIIT intervention on cognitive performance in older women.

Cardiorespiratory fitness (CRF) mitigates age-related decline in cognition and brain volume. Little is known, however, about the effects of high-intensity interval training (HIIT) on cognitive aging and the relationship between HIIT, cognition, hippocampal subfield volumes, and cerebral oxygen extraction fraction (OEF). Older sedentary women participated in an 8-week HIIT intervention. We conducted cognitive assessments, fitness assessments (VO2max), MRI scans: asymmetric spin echo oxygen extraction fraction (ASE-OEF), high-resolution multiple image co-registration and averaging (HR-MICRA) imaging, and transcranial Doppler ultrasonography before and after the intervention. VO2max increased from baseline (M = 19.36, SD = 2.84) to follow-up (M = 23.25, SD = 3.61), Z = - 2.93, p < .001, r = 0.63. Composite cognitive (Z = - 2.05, p = 0.041), language (Z = - 2.19, p = 0.028), and visuospatial memory (Z = - 2.22, p = 0.026), z-scores increased significantly. Hippocampal subfield volumes CA1 and CA3 dentate gyrus and subiculum decreased non-significantly (all p > 0.05); whereas a significant decrease in CA2 (Z = - 2.045, p = 0.041, r = 0.436) from baseline (M = 29.51; SD = 24.50) to follow-up (M = 24.50; SD = 13.38) was observed. Right hemisphere gray matter was correlated with language z-scores (p = 0.025; r = 0.679). The subiculum was correlated with attention (p = 0.047; r = 0.618) and verbal memory (p = 0.020; r = 0.700). The OEF and CBF were unchanged at follow-up (all p > .05). Although we observed cognitive improvements following 8 weeks of our HIIT intervention, they were not explained by hippocampal, OEF, or CBF changes.

Comparison of cerebral oxygen extraction fraction using ASE and TRUST methods in patients with sickle cell disease and healthy controls.

Abnormal oxygen extraction fraction (OEF), a putative biomarker of cerebral metabolic stress, may indicate compromised oxygen delivery and ischemic vulnerability in patients with sickle cell disease (SCD). Elevated OEF was observed at the tissue level across the brain using an asymmetric spin echo (ASE) MR method, while variable global OEFs were found from the superior sagittal sinus (SSS) using a T2-relaxation-under-spin-tagging (TRUST) MRI method with different calibration models. In this study, we aimed to compare the average ASE-OEF in the SSS drainage territory and TRUST-OEF in the SSS from the same SCD patients and healthy controls. 74 participants (SCD: N = 49; controls: N = 25) underwent brain MRI. TRUST-OEF was quantified using the Lu-bovine, Bush-HbA and Li-Bush-HbS models. ASE-OEF and TRUST-OEF were significantly associated in healthy controls after controlling for hematocrit using the Lu-bovine or the Bush-HbA model. However, no association was found between ASE-OEF and TRUST-OEF in patients with SCD using either the Bush-HbA or the Li-Bush-HbS model. Plausible explanations include a discordance between spatially volume-averaged oxygenation brain tissue and flow-weighted volume-averaged oxygenation in SSS or sub-optimal calibration in SCD. Further work is needed to refine and validate non-invasive MR OEF measurements in SCD.

An automatic pipeline for PET/MRI attenuation correction validation in the brain.

PURPOSE: Challenges in PET/MRI quantitative accuracy for neurological uses arise from PET attenuation correction accuracy. We proposed and evaluated an automatic pipeline to assess the quantitative accuracy of four MRI-derived PET AC methods using analytically simulated PET brain lesions and ROIs as ground truth for PET activity. METHODS: Our proposed pipeline, integrating a synthetic lesion insertion tool and the FreeSurfer neuroimaging framework, inserts simulated spherical and brain ROIs into PET projection space, reconstructing them via four PET MRAC techniques. Utilizing an 11-patient brain PET dataset, we compared the quantitative accuracy of four MRACs (DIXON, DIXONbone, UTE AC, and DL-DIXON) against the gold standard PET CTAC, evaluating MRAC to CTAC activity bias in spherical lesions and brain ROIs with and without background activity against original (lesion free) PET reconstructed images. RESULTS: The proposed pipeline yielded accurate results for spherical lesions and brain ROIs, adhering to the MRAC to CTAC pattern of original brain PET images. Among the MRAC methods, DIXON AC exhibited the highest bias, followed by UTE, DIXONBone, and DL-DIXON showing the least. DIXON, DIXONbone, UTE, and DL-DIXON showed MRAC to CTAC biases of - 5.41%, - 1.85%, - 2.74%, and 0.08% respectively for ROIs inserted in background activity; - 7.02%, - 2.46%, - 3.56%, and - 0.05% for lesion ROIs without background; and - 6.82%, - 2.08%, - 2.29%, and 0.22% for the original brain PET images' 16 FreeSurfer brain ROIs. CONCLUSION: The proposed pipeline delivers accurate results for synthetic spherical lesions and brain ROIs, with and without background activity consideration, enabling the evaluation of new attenuation correction approaches without utilizing measured PET emission data. Additionally, it offers a consistent method to generate realistic lesion ROIs, potentially applicable in assessing further PET correction techniques.

An automatic pipeline for PET/MRI attenuation correction validation in the brain.

Purpose: PET/MRI quantitative accuracy for neurological applications is challenging due to accuracy of the PET attenuation correction. In this work, we proposed and evaluated an automatic pipeline for assessing the quantitative accuracy of four different MRI = based attenuation correction (PET MRAC) approaches. Methods: The proposed pipeline consists of a synthetic lesion insertion tool and the FreeSurfer neuroimaging analysis framework. The synthetic lesion insertion tool is used to insert simulated spherical, and brain regions of interest (ROI) into the PET projection space and reconstructed with four different PET MRAC techniques, while FreeSurfer is used to generate brain ROIs from T1 weighted MRI image. Using a cohort of 11 patients' brain PET dataset, the quantitative accuracy of four MRAC(s), which are: DIXON AC, DIXONbone AC, UTE AC, and Deep learning trained with DIXON AC, named DL-DIXON AC, were compared to the PET-based CT attenuation correction (PET CTAC). MRAC to CTAC activity bias in spherical lesions and brain ROIs were reconstructed with and without background activity and compared to the original PET images. Results: The proposed pipeline provides accurate and consistent results for inserted spherical lesions and brain ROIs inserted with and without considering the background activity and following the same MRAC to CTAC pattern as the original brain PET images. As expected, the DIXON AC showed the highest bias; the second was for the UTE, then the DIXONBone, and the DL-DIXON with the lowest bias. For simulated ROIs inserted in the background activity, DIXON showed a -4.65% MRAC to CTAC bias, 0.06% for the DIXONbone, -1.70% for the UTE, and - 0.23% for the DL-DIXON. For lesion ROIs inserted without background activity, DIXON showed a -5.21%, -1% for the DIXONbone, -2.55% for the UTE, and - 0.52 for the DL-DIXON. For MRAC to CTAC bias calculated using the same 16 FreeSurfer brain ROIs in the original brain PET reconstructed images, a 6.87% was observed for the DIXON, -1.83% for DIXON bone, -3.01% for the UTE, and - 0.17% for the DL-DIXON. Conclusion: The proposed pipeline provides accurate and consistent results for synthetic spherical lesions and brain ROIs inserted with and without considering the background activity; hence a new attenuation correction approach can be evaluated without using measured PET emission data.

Increased theta-low gamma phase-amplitude coupling in resting electroencephalography after intermittent theta burst stimulation.

OBJECTIVE: Intermittent theta burst stimulation (iTBS) is based on the phase-amplitude coupling (PAC) pattern. We aimed to investigate the effect of iTBS on PAC in resting electroencephalography (EEG), which may provide insight into the underlying mechanism. METHODS: Twenty-one healthy volunteers were recruited and received both active and sham neuroimaging-guided iTBS on two separate days, which was precisely delivered to the right superior temporal gyrus. On each experimental day, resting EEG was recorded before and after stimulation for each participant. PACs across electrodes and frequency bands were calculated and compared to investigate the effect of iTBS. RESULTS: Theta (4-6 Hz) -low gamma (45-55 Hz) PAC over the stimulation site had a significant interaction effect, which increased after the active iTBS but did not differ after the sham iTBS. No significant interaction effect occurred in other cross-frequency couplings such as delta-low gamma, alpha-low gamma, delta-high gamma, theta-high gamma, or alpha-high gamma PAC in the region of interest. CONCLUSION: iTBS selectively modulated theta-low gamma PAC at the stimulation area, which exhibited both region- and frequency- specificity. This suggests that PAC may be a bridge connecting external neuromodulation to internal neuroplasticity.

Neuroinflammation and amyloid deposition in the progression of mixed Alzheimer and vascular dementia.

BACKGROUND: Alzheimer's disease (AD) and vascular contributions to cognitive impairment and dementia (VCID) pathologies coexist in patients with cognitive impairment. Abnormal amyloid beta (Aß) deposition is the hallmark pathologic biomarker for AD. Neuroinflammation may be a pathophysiological mechanism in both AD and VCID. In this study, we aimed to understand the role of neuroinflammation and Aß deposition in white matter hyperintensities (WMH) progression and cognitive decline over a decade in patients with mixed AD and VCID pathologies. METHODS: Twenty-four elderly participants (median [interquartile range] age 78 [64.8, 83] years old, 14 female) were recruited from the Knight Alzheimer Disease Research Center. 11C-PK11195 standard uptake value ratio (SUVR) and 11C-PiB mean cortical binding potential (MCBP) were used to evaluate neuroinflammation and Aß deposition in-vivo, respectively. Fluid-attenuated inversion recovery MR images were acquired to obtain baseline WMH volume and its progression over 11.5 years. Composite cognitive scores (global, processing speed and memory) were computed at baseline and follow-up over 7.5 years. Multiple linear regression models evaluated the association between PET biomarkers (11C-PK11195 SUVR and 11C-PiB MCBP) and baseline WMH volume and cognitive function. Moreover, linear mixed-effects models evaluated whether PET biomarkers predicted greater WMH progression or cognitive decline over a decade. RESULTS: Fifteen participants (62.5%) had mixed AD (positive PiB) and VCID (at least one vascular risk factor) pathologies. Elevated 11C-PK11195 SUVR, but not 11C-PiB MCBP, was associated with greater baseline WMH volume and predicted greater WMH progression. Elevated 11C-PiB MCBP was associated with baseline memory and global cognition. Elevated 11C-PK11195 SUVR and elevated 11C-PiB MCBP independently predicted greater global cognition and processing speed declines. No association was found between 11C-PK11195 SUVR and 11C-PiB MCBP. CONCLUSIONS: Neuroinflammation and Aß deposition may represent two distinct pathophysiological pathways, and both independently contributed to the progression of cognitive impairment in mixed AD and VCID pathologies. Neuroinflammation, but not Aß deposition, contributed to WMH volume and progression.

A Path to Qualification of PET/MRI Scanners for Multicenter Brain Imaging Studies: Evaluation of MRI-Based Attenuation Correction Methods Using a Patient Phantom.

PET/MRI scanners cannot be qualified in the manner adopted for hybrid PET/CT devices. The main hurdle with qualification in PET/MRI is that attenuation correction (AC) cannot be adequately measured in conventional PET phantoms because of the difficulty in converting the MR images of the physical structures (e.g., plastic) into electron density maps. Over the last decade, a plethora of novel MRI-based algorithms has been developed to more accurately derive the attenuation properties of the human head, including the skull. Although promising, none of these techniques has yet emerged as an optimal and universally adopted strategy for AC in PET/MRI. In this work, we propose a path for PET/MRI qualification for multicenter brain imaging studies. Specifically, our solution is to separate the head AC from the other factors that affect PET data quantification and use a patient as a phantom to assess the former. The emission data collected on the integrated PET/MRI scanner to be qualified should be reconstructed using both MRI- and CT-based AC methods, and whole-brain qualitative and quantitative (both voxelwise and regional) analyses should be performed. The MRI-based approach will be considered satisfactory if the PET quantification bias is within the acceptance criteria specified here. We have implemented this approach successfully across 2 PET/MRI scanner manufacturers at 2 sites.

The lateralization and reliability of spatial mismatch negativity elicited by auditory deviants with virtual spatial location.

Mismatch negativity (MMN) is an intensively studied event-related potential component that reflects pre-attentive auditory processing. Existing spatial MMN (sMMN) studies usually use loud-speakers in different locations or deliver sound with binaural localization cues through earphones to elicit MMN, which either was practically complicated or sounded unnatural to the subjects. In the present study, we generated head related transfer function (HRTF)-based spatial sounds and verified that the HRTF-based sounds retained the left and the right spatial localization cues. We further used them as deviants to elicit sMMN with conventional oddball paradigm. Results showed that sMMN was successfully elicited by the HRTF-based deviants in 18 of 21 healthy subjects in two separate sessions. Furthermore, the left deviants elicited higher sMMN amplitudes in the right hemisphere compared to the left hemisphere, while the right deviants elicited sMMN with similar amplitudes in both hemispheres, which supports a combination of contralateral and right-hemispheric dominance in spatial auditory information processing. In addition, the sMMN in response to the right deviants showed good test-retest reliability, while the sMMN in response to the left deviants had weak test-retest reliability. These findings implicate that HRTF-based sMMN could be a robust paradigm to investigate spatial localization and discrimination abilities.

Precise theta burst transcranial magnetic stimulation selectively reduced duration-related mismatch negativity.

BACKGROUND: Mismatch negativity (MMN) is a typical event-related potential component reflecting pre-attentive processing. MMN impairment, especially reduced duration-related MMN (dMMN), has been suggested as a potential predictive biomarker for the onset of schizophrenia. OBJECTIVE: This study attempts to manipulate specific MMN activities using advanced neuroimaging-guided intermittent theta-burst stimulations (iTBS), which will be helpful to uncover the sources of MMN generation and contribute to the development of new clinical treatments. METHODS: Twenty-four healthy volunteers were recruited and participated two-session modulations consisting of active and sham iTBS. ITBS was precisely delivered over individual right posterior superior temporal cortex (pSTG). Before and after each iTBS session, two MMN components evoked by duration and frequency deviants were quantified respectively. RESULTS: A significant interaction of time and iTBS was observed on dMMN amplitudes, but not frequency-related MMN amplitudes. dMMN only decreased after active precise iTBS intervention, but did not after sham iTBS. The post effect of iTBS on dMMN was found in 16 of 20 subjects, suggesting a robust effect even at individual level. Furthermore, sLORETA analysis showed that the lateralization of STG activation was reversed after the active iTBS. CONCLUSIONS: We applied a precise strategy for neuroimaging-guided iTBS modulation over the right pSTG, which is promising in selectively modulating MMN for specific deviants.