Privacy-friendly mobility analytics using aggregate location data

Brain Magnetic Resonance Imaging Findings in Children and Young Adults With CKD

OBJECTIVES/GOALS: Klotho is a protein linked to improved cognition in aging adults. A specific KLOTHO gene variant, KL-VS, increases circulating levels of Klotho. The current study aims to identify if the KL-VS haplotype and Klotho levels are associated with improved neurocognition in pediatric brain tumor survivors. METHODS/STUDY POPULATION: We are actively accruing pediatric brain tumor patients at UCSF alongside an existing multi-institutional cohort study investigating radiation-induced vasculopathies and cognitive outcomes in this population. Normal controls are being enrolled in parallel. Each patient undergoes: 1) single nucleotide polymorphism genotyping to identify KL-VS haplotype status, 2) enzyme-linked immunosorbent assays to measure circulating Klotho, 3) neurocognitive assessments with a computer-based, validated Cogstate battery, and 4) brain volume and white matter lesion segmentation analyses using MRI sequences obtained as part of routine care. RESULTS/ANTICIPATED RESULTS: Genotyping has been performed on 99 enrolled patients. KL-VS heterozygosity was seen in 22.7% of patients. To date, KL-VS status is not associated with neurocognitive outcomes at baseline or Year 1 testing. Association between KL-VS status, circulating Klotho levels, neurocognitive outcomes, brain volume and white matter lesion segmentation analyses is ongoing. We hypothesize that elevated Klotho levels will be associated with improved neurocognition, increased brain volumes in regions of interest and decreased white matter lesion volumes. DISCUSSION/SIGNIFICANCE OF IMPACT: If circulating Klotho leads to improved neurocognition in pediatric brain tumor survivors, Klotho levels might serve as a prognostic biomarker. Furthermore, as Klotho is being investigated for therapeutic indications, it may represent an intervention to prevent cognitive deficits in these patients.

To prospectively determine clinical value of dynamic contrast material-enhanced magnetic resonance (MR) imaging in the evaluation of disease activity in perianal Crohn disease (CD). Patients provided written informed consent. Study approval was waived by an institutional review board. Thirty-three patients with perianal CD underwent pelvic MR imaging; 17 were male (mean age, 37.4 years +/- 10.8 [standard deviation]; age range, 18-54 years) and 16 were female (mean age, 32.0 years +/- 8.3; age range, 16-43 years). Dynamic contrast-enhanced MR imaging was performed; time-intensity curves (TICs) were obtained. Each pixel was classified as one of six predefined TIC shape types. For each MR imaging examination, a region of interest (ROI) was drawn around the fistula on the single section corresponding to the most extensive and most hyperintense lesion; maximum enhancement (ME), slope of enhancement, and TIC shapes were calculated. Absolute and relative numbers of pixels for each curve type were calculated in a two-dimensional ROI. These results were compared with Perianal Disease Activity Index (PDAI), C-reactive protein (CRP) level, an MR imaging-based severity score, and clinical outcome. A Spearman rank correlation test was used to calculate correlation coefficients between dynamic contrast-enhanced MR imaging parameters and reference parameters. A Mann-Whitney U test was used to calculate differences in dynamic contrast-enhanced MR imaging parameters between predefined groups of patients. Significant correlations were found between the absolute amounts of the TIC shape types and PDAI and between ROI volume and PDAI. The ratio of quickly enhancing versus slowly enhancing pixels correlated with higher MR imaging scores as did the ROI volume. The absolute amounts of pixels displaying TIC types 2, 3, 4, and 5 correlated significantly with MR imaging score. CRP level showed a significant correlation with mean ME. Larger numbers of quickly enhancing pixels were observed in patients who needed medication changes or developed new abscesses during follow-up. Dynamic contrast-enhanced MR imaging can help determine disease activity in perianal CD and might be helpful in selecting a subpopulation of patients who should be monitored more closely for development of more extensive disease.

A functional gradient has been proposed across the medial temporal lobes (MTL) such that the anterior MTL is thought to support processing of individual items (e.g., item memory and complex object perception), whereas the posterior MTL is thought to support item-context retrieval (e.g., source memory). Whereas functional imaging studies have provided evidence supporting this anatomical organization, results from structural analyses remain inconclusive. The current study examined the relationship between volume of MTL regions of interest (ROIs), and performance on a source memory task and a fine-grain complex object perception task, in healthy young adults (mean age = 21.5, range = 18-29). Using a semiautomated procedure, we segmented the parahippocampal and perirhinal cortices (PHC, PRC), posteromedial and anterolateral entorhinal cortices (pmERC, alERC), and posterior and anterior hippocampus (postHC, antHC) on high-resolution T2-weighted MRIs. Regional volumes were computed as proportions of intracranial volume, and as posterior-anterior volumetric ratios (PHC:PRC, pmERC:alERC, postHC:antHC). Partial-least squares regressions were applied to predict source and item memory, and perceptual discrimination accuracy, based on ROI and ratio volumes. In our ROI regressions, we found that postHC volume was positively correlated with a latent factor predicting source memory, and PRC and antHC volumes were negatively correlated to this latent factor. In our ratio regressions, we observed an effect relating the posterior-anterior distribution of gray matter across the MTL with source memory. Our results demonstrate differential associations between anterior and posterior MTL and source memory performance. Findings from this study highlight the importance of considering patterns of structure-behavior associations in the neurobiology of episodic memory.

A numerical study is conducted of the interaction between the leading oblique detonation wave stabilized by a wedge ramp of finite length and the expansion waves generated by the sudden deflection of the wedge surface. The two-dimensional Euler equations for a reactive gas mixture are solved using a cell-centered finite volume technique on unstructured meshes. Mesh refinement and coarsening techniques are used to concentrate the computational volumes in regions of interest of the flowfield. A detailed chemical kinetic mechanism is employed to describe the combustion process in the mixtures of hydrogen and air considered. Computational results show that a Chapman‐ Jouguet oblique detonation wave is obtained for intermediate wedge angles within the range of stable detonations. For wedge angles near to the maximum angle that supports stable oblique detonations, decoupling of the oblique detonation wave is observed.

Location data is frequently collected from populations and shared in aggregate form to guide policy and decision making. However, the prevalence of aggregated data also raises the privacy concern of membership inference attacks (MIAs). MIAs infer whether an individual's data contributed to the aggregate release. Although effective MIAs have been developed for aggregate location data, these require access to an extensive auxiliary dataset of individual traces over the same locations, which are collected from a similar population. This assumption is often impractical given common privacy practices surrounding location data. To measure the risk of an MIA performed by a realistic adversary, we develop the first Zero Auxiliary Knowledge (ZK) MIA on aggregate location data, which eliminates the need for an auxiliary dataset of real individual traces. Instead, we develop a novel synthetic approach, such that suitable synthetic traces are generated from the released aggregate. We also develop methods to correct for bias and noise, to show that our synthetic-based attack is still applicable when privacy mechanisms are applied prior to release. Using two large-scale location datasets, we demonstrate that our ZK MIA matches the state-of-the-art Knock-Knock (KK) MIA across a wide range of settings, including popular implementations of differential privacy (DP) and suppression of small counts. Furthermore, we show that ZK MIA remains highly effective even when the adversary only knows a small fraction (10%) of their target's location history. This demonstrates that effective MIAs can be performed by realistic adversaries, highlighting the need for strong DP protection.

An investigation of statistical power for continuous arterial spin labeling imaging at 1.5 T

Chronotype refers to individual preference in circadian cycles and is associated with psychiatric problems. It is mainly classified into early (those who prefer to be active in the morning and sleep and wake up early) and late (those who prefer to be active in the evening and sleep and wake up late) chronotypes. Although previous research has demonstrated associations between chronotype and cognitive function and brain structure in adults, little is known regarding these associations in children. Here, we aimed to investigate the relationship between chronotype and cognitive function in children. Moreover, based on the significant association between chronotype and specific cognitive functions, we extracted regions-of-interest (ROI) and examined the association between chronotype and ROI volumes. Data from 4,493 children (mean age of 143.06 months) from the Adolescent Brain Cognitive Development Study were obtained, wherein chronotype (mid-sleep time on free days corrected for sleep debt on school days) was assessed by the Munich Chronotype Questionnaire. Subsequently, the associations between chronotype, cognitive function, and ROI volumes were evaluated using linear mixed-effects models. Behaviorally, chronotype was negatively associated with vocabulary knowledge, reading skills, and episodic memory performance. Based on these associations, the ROI analysis focused on language-related and episodic memory-related areas revealed a negative association between chronotype and left precentral gyrus and right posterior cingulate cortex volumes. Furthermore, the precentral gyrus volume was positively associated with vocabulary knowledge and reading skills, while the posterior cingulate cortex volume was positively associated with episodic memory performance. These results suggest that children with late chronotype have lower language comprehension and episodic memory and smaller brain volumes in the left precentral gyrus and right posterior cingulate cortex associated with these cognitive functions.

BackgroundAlzheimer’s disease (AD) is associated with widespread reductions in cerebral blood flow (CBF), but the precise link between AD pathology and CBF is unclear. Hypoperfusion is a late‐stage event associated with tau, but not amyloid, pathology. We assessed whether baseline levels of cerebrospinal fluid (CSF)‐measured core AD pathology [β‐amyloid42 (Aβ42), phosphorylated tau (p‐tau)] related to CBF changes over a 7‐year period and if associations differed by baseline cognitive status.MethodsVanderbilt Memory and Aging Project participants free of clinical dementia or stroke at enrollment [n = 153, 72±6 years, 37% mild cognitive impairment (MCI), 33% female] underwent lumbar puncture to obtain CSF at study entry and serial multimodal 3T brain magnetic resonance imaging over a 7‐year follow‐up period (mean = 5.4 years). Enzyme‐linked immunosorbent assays quantified CSF Aβ42 and p‐tau181. Pseudo‐continuous arterial spin‐labeling captured CBF in total and lobar grey matter regions of interest (ROI). Linear mixed‐effects models related biomarker x time to longitudinal CBF variables adjusting for baseline age and cognitive status, sex, race/ethnicity, education, Framingham Stroke Risk Profile (minus age), ROI volume, APOE‐ε4 status, and time. Models were repeated testing a cognitive status x biomarker x time interaction term.ResultsIn main effect models, neither CSF Aβ42 (p‐values>0.31) nor p‐tau (p‐values>0.28) related to CBF outcomes. Aβ42 did not interact with cognitive status on CBF outcomes (p‐values>0.37), but p‐tau interacted with baseline cognitive status on longitudinal frontal, temporal, parietal, occipital, and total grey matter CBF trajectories (p‐values<0.005). Among cognitively unimpaired participants, higher p‐tau was associated with slower reductions in frontal (β = 0.01, p = 0.03), parietal (β = 0.01, p = 0.04), occipital (β = 0.01, p = 0.02), and total grey matter CBF (β = 0.01, p = 0.04). Among individuals with MCI, higher p‐tau was associated with faster reductions in frontal (β = ‐0.03, p = 0.008), temporal (β = ‐0.02, p = 0.01), parietal (β = ‐0.04, p = 0.002), occipital (β = ‐0.03, p = 0.0004), and total grey matter CBF (β = ‐0.03, p = 0.005).ConclusionsResults linking CSF p‐tau to CBF decline in MCI support the theory that grey matter hypoperfusion is a later‐stage event tied to neurofibrillary tangle pathology. Associations between higher p‐tau and slower rates of CBF decline among cognitively unimpaired individuals warrants further investigation. Funding: R01‐AG034962, K24‐AG046373, F31‐AG079640, T32‐AG058524, P20‐AG068082

Background: Elevated carotid intima-media thickness (cIMT) and presence of carotid plaque are markers of atherosclerosis, which may be linked to structural brain injury. Using data from the ARIC-NCS cohort, we tested the hypotheses that greater cIMT and presence of carotid plaque at late midlife were each associated with presence of infarcts (lacunar and cortical) and cerebral microbleeds (lobar and subcortical), greater white matter hyperintensity (WMH) volume, and smaller regional brain volumes in late life. Methods: We included 1,795 ARIC participants who had carotid ultrasounds in 1990-92 and brain MRI scans in 2011-13. Weighted linear regression was used for continuous (brain volume) outcomes, and weighted logistic regression for dichotomous outcomes (infarcts and cerebral microbleeds). Results: Participants were aged 57±6 years at baseline, 57% were female, and 23% were Black. Mean ± SD follow-up was 21±1 years. After multivariable adjustments, the highest quintile of cIMT in late midlife was associated with smaller deep grey matter volume (β [95% CI]: -0.11 [-0.22, -0.01]) and cortical volume in a temporal-parietal meta region of interest (ROI) (β [95% CI]: -0.10 [-0.20, -0.01]) in late life (Table). At baseline, 473 (26%) participants had carotid plaque present on their carotid ultrasound. Similar to cIMT, those with carotid plaque had smaller regional brain volumes than those without carotid plaque present (βs [95% CIs]: -0.05 [-0.12, 0.03] and -0.06 [-0.13, 0.01] for deep grey matter and temporal-parietal meta ROI volumes, respectively). No significant relations were observed with WMH volume, or the presence of any infarcts or cerebral microbleeds. Conclusion: Greater midlife cIMT and presence of carotid plaque were associated with smaller deep grey matter volume and cortical volume in a temporal-parietal meta ROI in later life. Contrary to our hypothesis, associations between midlife subclinical atherosclerosis and markers of vascular brain injury in later life were null.

Background/Aims Sleep problems are common in fibromyalgia (FM) and chronic widespread pain (CWP), and may result in similar structural brain changes as those seen in chronic pain. However, the exact neural mechanisms are not fully understood. Studies to date are relatively small and yield conflicting results. The aim of this study is to investigate structural differences in brain regions of interest (ROI) in participants with FM or CWP from UK Biobank, a large-scale biomedical database, compared with healthy controls. We also explore the potential interaction of sleep on the volume of these ROIs, with the hypothesis that low sleep duration with chronic pain amplifies the atrophy seen. Methods Using T1 structural brain imaging data from UK Biobank, we defined cases as those experiencing CWP for 3+ months at the imaging visit or with a diagnosis of FM. Age- and sex-matched controls did not report any pain or chronic pain diagnoses at the imaging visit. Low sleep duration was defined as less than 7 hours of sleep in a 24-hour period in line with the recommended sleep duration of 7-9 hours. 22 brain ROIs were selected based on the existing literature. Pre-processed image-derived phenotype data for grey matter volumes were compared in brain ROIs using 2-way independent ANOVAs in SPSS. Results 429 cases and 429 controls were included: (n = 664/858 [77.4%] female; mean [SD] age 62.0 ± 7.4). The volume of the grey matter in the bilateral insular cortices and anterior parahippocampal gyri, and the volume of the right pallidum, were all significantly smaller in cases compared to controls, p < 0.05. In addition to the significant difference between the controls and cases in the right anterior parahippocampal gyrus, the interaction effect of sleep duration on pain was also significant, p < 0.05. Even though there were no significant differences between the controls and cases in the grey matter volumes of the left posterior division of the cingulate gyrus, bilateral superior frontal gyri, right middle frontal gyrus, left posterior division of the parahippocampal gyrus, right precuneus cortex and left inferior frontal gyrus, pars opercularis, and the volume of the right thalamus, there was a significant interaction effect of sleep duration and pain status on their volumes, p < 0.05. Conclusion This large-scale study confirms differences in grey matter volumes in key brain ROIs in those with chronic pain, highlighting neural changes in individuals suffering from CWP or FM, and a novel interaction of sleep on these areas. Perhaps these changes are due to neurotoxic effects of low sleep duration. Alternatively, low sleep duration and chronic pain may be caused by these structural differences. Evidently, further research is still needed to identify the directionality of these changes and their functional implications through longitudinal functional brain imaging studies, to identify the underlying cause of these structural differences. Disclosure N.W.I.L. Silva: None. A. Soni: None. V. Wanigasekera: None. S. Zhang: None. A.J.W. Wall: None. I. Tracey: None.

Transportation projects require an enormous amount of capital investment resulting from their tremendous size, complexity, and risk. Due to the limitation of public finances, the private sector is invited to participate in transportation project development. The private sector can entirely or partially invest in transportation projects in the form of Public-Private Partnership (PPP) scheme, which has been an attractive option for several developing countries, including Vietnam. There are many factors affecting the success of PPP projects. The accurate prediction of traffic volume is considered one of the key success factors of PPP transportation projects. However, only few research works investigated how to predict traffic volume over a long period of time. Moreover, conventional traffic volume forecasting methods are usually based on deterministic models which predict a single value of traffic volume but do not consider risk and uncertainty. This knowledge gap makes it difficult for concessionaires to estimate PPP transportation project revenues accurately. The objective of this paper is to develop a probabilistic traffic volume prediction model. First, traffic volumes were estimated following the Geometric Brownian Motion (GBM) process. Monte Carlo technique is then applied to simulate different scenarios. The results show that this stochastic approach can systematically analyze variations in the traffic volume and yield more reliable estimates for PPP projects.

Proactive traffic management is a subset of smart mobility applications in which traffic control strategies are implemented in advance to respond to anticipated roadway conditions. Predicted traffic flows are a key input to proactive traffic control systems, such as proactive freeway ramp metering, proactive variable speed limits, and proactive incident management systems. This paper proposes a machine learning approach for short-term traffic flow prediction where prevailing conditions, such as the traffic volume, speed, and occupancy of roadway segments, are used to predict traffic flow in short-term intervals. Four categories of predictive methods for traffic flow prediction were investigated: deep neural networks, a distributed random forest, a gradient boosting machine, and a generalized linear model. Data from Interstate 64 in St. Louis, Missouri, in the United States were used to calibrate and evaluate the models. The results obtained by the four predictive methods were very similar, with the distributed random forest model slightly outperforming the models obtained by the other three methods. The case study showed that the inclusion of traffic flow, speed, occupancy, and time of day in the traffic prediction process reduces the traffic prediction error. However, the two-sample Kolmogorov-Smirnov test did not show a statistically significant benefit from the inclusion of upstream traffic data in the distributed random forest, and gradient boosting machine models.

Objective Volume reduction and structural abnormality is the most replicated finding in neuroimaging studies of Alzheimer's disease (AD). Amnestic mild cognitive impairment (aMCI) is the early stage of AD development. Thus, it is necessary to investigate the link between atrophy of regions of interest (ROIs) in medial temporal lobe, the variation trend of ROI densities and volumes among patients with cognitive impairment, and the distribution characteristics of ROIs in the aMCI group, Alzheimer's disease (AD) group, and normal control (NC) group. Methods 30 patients with aMCI, 16 patients with AD, and 30 NC are recruited; magnetic resonance imaging (MRI) brain scans are conducted. Voxel-based morphometry was employed to conduct the quantitative measurement of gray matter densities of the hippocampus, amygdala, entorhinal cortex, and mammillary body (MB). FreeSurfer was utilized to automatically segment the hippocampus into 21 subregions and the amygdala into 9 subregions. Then, their subregion volumes and total volume were calculated. Finally, the ANOVA and multiple comparisons were performed on the above-mentioned data from these three groups. Results AD had lower GM densities than MCI, and MCI had lower GM densities than NC, but not all of the differences were statistically significant. In the comparisons of AD-aMCI-NC, AD-aMCI, and AD-NC, the hippocampus, amygdala, and entorhinal cortex showed differences in the gray matter densities (p < 0.05); the differences of mammillary body densities were not significant in the random comparison between these three groups (p > 0.05). The hippocampus densities and volumes of the subjects from the aMCI group and the AD group were bilaterally symmetric. The gray matter densities of the right side of the entorhinal cortex inside each group and the hippocampus from the NC group were higher than those of the left side (p < 0.05), and the gray matter densities of the amygdala and mammillary body were bilaterally symmetric in the three groups (p > 0.05). There were no gender differences of four ROIs in the AD, aMCI, and NC groups (p > 0.05). The volume differences of the hippocampus presubiculum-body and parasubiculum manifest no statistical significance (p > 0.05) in the random comparison between these three groups. Volume differences of the left amygdala basal nucleus, the left lateral nucleus, the left cortical amygdala transitional area, the left paravamnion nucleus, and bilateral hippocampal amygdala transition area (HATA) had statistical differences only between the AD group and the NC group (p < 0.05). Conclusion Structural defects of medial temporal lobe subfields were revealed in the aMCI and AD groups. Decreased gray matter densities of the hippocampus, entorhinal cortex, and amygdala could distinguish patients with early stage of AD between aMCI and NC. Volume decline of the hippocampus and amygdala subfields could only distinguish AD between NC.

Proton computed tomography (pCT) has high accuracy and dose efficiency in producing spatial maps of the relative stopping power (RSP) required for treatment planning in proton therapy. With fluence-modulated pCT (FMpCT), prescribed noise distributions can be achieved, which allows to decrease imaging dose by employing object-specific dynamically modulated fluence during the acquisition.For FMpCT acquisitions we divide the image into region-of-interest (ROI) and non-ROI volumes. In proton therapy, the ROI volume would encompass all treatment beams. An optimization algorithm then calculates dynamically modulated fluence that achieves low prescribed noise inside the ROI and high prescribed noise elsewhere. It also produces a planned noise distribution, which is the expected noise map for that fluence, as calculated with a Monte Carlo simulation. The optimized fluence can be achieved by acquiring pCT images with grids of intensity modulated pencil beams. In this work, we interfaced the control system of a clinical proton beam line to deliver the optimized fluence. Using three phantoms we acquired images with uniform fluence, with a constant noise prescription, and with an FMpCT task. Image noise distributions as well as fluence maps were compared to the corresponding planned distributions as well as to the prescription. Furthermore, we propose a correction method that removes image artifacts stemming from the acquisition with pencil beams having a spatially varying energy distribution that is not seen in clinical operation. RSP accuracy of FMpCT scans was compared to uniform scans and was found to be comparable to standard pCT scans.While we identified technical improvements for future experimental acquisitions, in particular related to an unexpected pencil beam size reduction and a misalignment of the fluence pattern, agreement with the planned noise was satisfactory and we conclude that FMpCT optimized for specific image noise prescriptions is experimentally feasible.