Total Voxels Research Articles

Uncovering the hidden effects of repetitive subconcussive head impact exposure: A mega-analytic approach characterizing seasonal brain microstructural changes in contact and collision sports athletes.

Repetitive subconcussive head impacts (RSHI) are believed to induce sub-clinical brain injuries, potentially resulting in cumulative, long-term brain alterations. This study explores patterns of longitudinal brain white matter changes across sports with RSHI-exposure. A systematic literature search identified 22 datasets with longitudinal diffusion magnetic resonance imaging data. Four datasets were centrally pooled to perform uniform quality control and data preprocessing. A total of 131 non-concussed active athletes (American football, rugby, ice hockey; mean age: 20.06 ± 2.06 years) with baseline and post-season data were included. Nonparametric permutation inference (one-sample t tests, one-sided) was applied to analyze the difference maps of multiple diffusion parameters. The analyses revealed widespread lateralized patterns of sports-season-related increases and decreases in mean diffusivity (MD), radial diffusivity (RD), and axial diffusivity (AD) across spatially distinct white matter regions. Increases were shown across one MD-cluster (3195 voxels; mean change: 2.34%), one AD-cluster (5740 voxels; mean change: 1.75%), and three RD-clusters (817 total voxels; mean change: 3.11 to 4.70%). Decreases were shown across two MD-clusters (1637 total voxels; mean change: -1.43 to -1.48%), two RD-clusters (1240 total voxels; mean change: -1.92 to -1.93%), and one AD-cluster (724 voxels; mean change: -1.28%). The resulting pattern implies the presence of strain-induced injuries in central and brainstem regions, with comparatively milder physical exercise-induced effects across frontal and superior regions of the left hemisphere, which need further investigation. This article highlights key considerations that need to be addressed in future work to enhance our understanding of the nature of observed white matter changes, improve the comparability of findings across studies, and promote data pooling initiatives to allow more detailed investigations (e.g., exploring sex- and sport-specific effects).

18F-Flortaucipir (AV1451) imaging identifies grey matter atrophy in retired athletes

BackgroundThe long-term consequences of concussions may include pathological neurodegeneration as seen in Alzheimer’s disease (AD) and chronic traumatic encephalopathy (CTE). Tau-PET showed promise as a method to detect tau pathology of CTE, but more studies are neededObjectiveThis study aimed (1) to assess the association of imaging evidence of tau pathology with brain volumes in retired athletes and (2) to examine the relationship between tau-PET and neuropsychological functioning.MethodsFormer contact sport athletes were recruited through the Canadian Football League Alumni Association or the Canadian Concussion Centre clinic. Athletes completed MRI, [18F]flortaucipir tau-PET, and a neuropsychological battery. Memory composite was created by averaging the Rey Auditory Verbal Learning Test and Rey Visual Design Learning Test z-scores. Grey matter (GM) volumes were age/intracranial volume corrected using normal control MRIs. Tau-PET % positivity in GM was calculated as the number of positive voxels (≥ 1.3 standardized uptake value ratio (SUVR)/total voxels).Results47 retired contact sport athletes negative for AD (age:51 ± 14; concussions/athlete:15 ± 2) and 54 normal controls (age:50 ± 13) were included. Tau-PET positive voxels had significantly lower GM volumes, compared to tau-PET negative voxels (− 0.37 ± 0.41 vs. − 0.31 ± 0.37, paired p = .006). There was a significant relationship between GM tau-PET % positivity and memory composite score (r = − .366, p = .02), controlled for age, PET scanner, and PET scan duration. There was no relationship between tau-PET measures and concussion number, or years of sport played.ConclusionA higher tau-PET signal was associated with reduced GM volumes and lower memory scores. Tau-PET may be useful for identifying those at risk for neurodegeneration.

Quantifying lung fissure integrity using a three-dimensional patch-based convolutional neural network on CT images for emphysema treatment planning.

Evaluation of lung fissure integrity is required to determine whether emphysema patients have complete fissures and are candidates for endobronchial valve (EBV) therapy. We propose a deep learning (DL) approach to segment fissures using a three-dimensional patch-based convolutional neural network (CNN) and quantitatively assess fissure integrity on CT to evaluate it in subjects with severe emphysema. From an anonymized image database of patients with severe emphysema, 129 CT scans were used. Lung lobe segmentations were performed to identify lobar regions, and the boundaries among these regions were used to construct approximate interlobar regions of interest (ROIs). The interlobar ROIs were annotated by expert image analysts to identify voxels where the fissure was present and create a reference ROI that excluded non-fissure voxels (where the fissure is incomplete). A CNN configured by nnU-Net was trained using 86 CT scans and their corresponding reference ROIs to segment the ROIs of left oblique fissure (LOF), right oblique fissure (ROF), and right horizontal fissure (RHF). For an independent test set of 43 cases, fissure integrity was quantified by mapping the segmented fissure ROI along the interlobar ROI. A fissure integrity score (FIS) was then calculated as the percentage of labeled fissure voxels divided by total voxels in the interlobar ROI. Predicted FIS (p-FIS) was quantified from the CNN output, and statistical analyses were performed comparing p-FIS and reference FIS (r-FIS). The absolute percent error mean (±SD) between r-FIS and p-FIS for the test set was 4.0% (), 6.0% (), and 12.2% () for the LOF, ROF, and RHF, respectively. A DL approach was developed to segment lung fissures on CT images and accurately quantify FIS. It has potential to assist in the identification of emphysema patients who would benefit from EBV treatment.

Validation of high resolution subcortical MRI signal using voxel to voxel histology to neuroimaging correlation in AD continuum: advancing neuroimaging biomarker tools for early AD staging

AbstractBackgroundStudies suggest a decrease in LC volume by 8% each Braak stage, starting at Braak 0‐1, while in normal aging the LC volume remains intact. Thus, LC volumetry is an attractive potential biomarker to monitor AD progression from precognitive stages. However, precise LC volumetry neuroimaging is challenging due to the LC’s small size and location in the brainstem that’s prone to artifacts. To validate the precision/utility of neuroimaging sequences designed for LC volumetry, we developed a pipeline allowing voxel‐to‐voxel histology to neuroimaging comparisons. We tested this pipeline to analyze the neuro‐histological counterpart of an LC volume obtained from manual segmentation from a MRI neuromelanin(NM)sensitive GRE image.MethodWhole brain high resolution(0.7mm)3D‐T1 images and 2D‐2mm thick NM‐sensitive axial GRE images of the brainstem were obtained postmortem in situ before brain removal from a single subject using a 7T MRI(Siemens Magnetom). We combined celloidin‐based histological processing and 3D reconstructions of the histology volume at microscopic resolution which was then aligned to T1. We used structures visible in structural MRI(T1)(brainstem surface, red nucleus, and 4th ventricle) to calculate Dice coefficient to measure registration accuracy. Next, we performed an overlap percentage analysis to investigate how well GRE sequence detected LC borders compared to histology.ResultWe ran Dice coefficient analysis for brainstem structures to measure histology to MRI registration accuracy(Table1). The LC GRE to histology masks overlap percentage was 35.29% and 10.77% compared to the MRI and histology mask total voxels(voxel size = 0.34mm3). The overlap percentage obtained for the rostral and middle LC thirds ranged from 9.43% to 75%when compared to total voxels in MRI/histology masks, while the caudal third had zero overlap(Table2,Fig.1).ConclusionOur pipeline enables voxel‐to‐voxel correlation analysis between histology and high resolution LC MRIs for neuroimaging validation. The caudal third of the LC in our case has shown a clear lack of overlap potentially due to lack of signal specificity in MRI when compared to the gold standard that is histology. Current analysis of another five other cases is ongoing. Validation of LC MRI signal will inform interpretations of signal results in living patients with the potential to act as a diagnostic tool for ADRD.

P01.11.A MECHANISMS OF LANGUAGE PLASTICITY IN PATIENTS HARBORING LEFT HEMISPHERE GLIOMAS

Abstract BACKGROUND Infiltrating gliomas are slow growing intrinsic brain lesions that have been shown to induce functional reorganization, although the mechanisms by which this occurs are not well understood. The present study examines glioma-induced neuroplasticity of language networks to investigate this phenomenon. MATERIAL AND METHODS 117 patients’ fMRI scans were analyzed from patients harboring gliomas whom completed 1-5 fMRI language tasks. Data was analyzed via two methods: cluster-based and cortical/subcortical mask-based. Right hemisphere gliomas (RHG) (n=40) were first analyzed to establish a reliable control group by determining statistically significant voxels across all five language tasks (LT), yielding a core conserved network. The language network induced by left hemisphere gliomas (LHG) (n=77) was then directly compared to the RHG language network to examine language plasticity. RESULTS LHG induced global reorganization of the canonical language network, including preserved homologous and novel left hemisphere (LH) and right hemisphere (RH) activation clusters. In each of the five LT, the LHG group showed activation of fewer total suprathreshold voxels in both the cluster (3 fewer clusters and 38,892 fewer active voxels/task) and mask-based (8 fewer active masks/task and 91,965 fewer active voxels/task) analyses while having a higher peak intensity (5.4 total increase in zmax) within the activated clusters. In addition, a generalized pattern of increased RH recruitment was seen in LHG patients, with 87% (137/157) of non-zero masks shifting their laterality index (LI) to the right hemisphere. These data suggest that the reorganized LHG-induced language network functions through less nodes along with a rightward shift of language lateralization. In addition, a subgroup analysis of frontal (n=16) and temporal (n=32) LHG revealed that frontal tumors promote increased RH activation relative to temporal tumors, with 94% (32/34) of non-zero masks shifting their LI to the RH compared to 68% (23/34), respectively. Frontal tumors also preferentially recruit the contralateral (RH) frontal lobe and ipsilateral (LH) temporal lobe, while temporal gliomas preferentially recruit the ipsilateral frontal lobe. CONCLUSION Infiltrating tumors within the left hemisphere promote language plasticity via consolidation of the classical LH language network along with recruitment of previously silent, novel language areas in the right and left hemispheres. This improved understanding of patterns of language plasticity may aid in more accurate pre-surgical risk stratification and counseling, therapeutic targeting for post-surgical recovery, and intelligent “pre-habilitation” strategies to maximize neuroplasticity prior to tumor resection.

Abstract TP90: Lesion Topography Role On Functional Outcome Of Stroke Patients

Background: The anatomical site of ischemia in stroke patients might impact functional outcome and can be a relevant factor for prognosis and clinical decision. Methods: We included patients from the MR CLEAN stroke trial population who had an adequate 5-9-day control CT scan. The images were segmented and registered over two neuroanatomical atlases. Each anatomical region was classified as affected or spared based on the total amount of voxels in each region interpreted as ischemic on CT imaging. Analysis were conducted with a threshold level of 20% and 50% of the total voxels as a binary classification criteria for considering a brain area as affected by the ischemic core. Univariate and multivariate analysis were conducted to assess for relationships between specific brain regions involvement and worse functional outcome (defined by the 90-day mRS). Results: A total of 356 patients was included. The left dorsal striatum and left insular ribbon were significantly associated with worse outcomes after correction for the total lesion volume (p < 0,05). On a binary logistic model, no areas were significantly associated with worse outcome when a threshold of 20% was used. Both left and right dorsal striatum’s were significantly associated with worse outcome after adjustment for volume when a threshold of 50% was used. Conclusion: Some areas of the brain appear to have a higher impact on clinical outcomes for stroke patients. Areas related to motor function were strongly associated with the 90-day mRS when compared to other brain areas. The impact of a brain region on outcome also seems to depend on which hemisphere has been affected and on the extent to which the region that has ben involved in the ischemia.

Intracranial computed tomography histogram analysis detects changes in the setting of elevated intracranial pressure and normal imaging.

Patients with idiopathic intracranial hypertension (IIH) have elevated intracranial pressure (ICP) of unclear etiology. This study evaluated the ability of quantitative intracranial Hounsfield unit (HU) histogram analysis to detect pathophysiological changes from elevated ICP in the setting of a normal head CT. Retrospective analysis of non-contrast-enhanced head CT images of IIH patients and matched controls. Following skull stripping, total intracranial CT voxels within the range of 0-70 HU were divided into seven 10 HU bins. A measurement of total intracranial HU was also calculated for each patient. Imaging studies for IIH patients were reviewed for features of IIH including transverse sinus stenosis (TSS). Histogram measures were compared between IIH and control groups and correlated with imaging and clinical data. Fourteen IIH patients with CSF opening pressure ≥25cm water, and 31 age-, sex-, and ethnicity-matched controls were included. Compared to controls, IIH patients had a significantly greater proportion of voxels in the 40-50, 50-60, and 60-70 HU bins (p = 0.003, 0.001, and 0.003, respectively) but similar proportion in the 0-10 HU range. Severity of TSS significantly correlated with total intracranial HU measures. 50-60 HU and 60-70 HU bins demonstrated high AUCs of 0.81 and 0.80, respectively, in differentiating IIH from normal status. Idiopathic intracranial hypertension patients have a greater proportion of high intracranial HU voxels representing blood volume, which may be explained by TSS causing venous congestion. The pattern provides further insights into the pathophysiology of IIH and may be useful for detecting elevated ICP in the setting of normal head CT imaging.

Abstract WP12: Evolution Of Regional Flair White Matter Hyperintensities After Ischemic Stroke: A Voxel-wise Approach

Background: White matter (WM) degeneration after ischemic stroke may herald post-stroke cognitive decline. FLAIR WM hyperintensity (WMH) reflects tissue injury and accumulates around the infarct and in remote regions. But WMH volumetry is insensitive to regional progression. We compared WMH volumetry to a voxel-wise (VW) approach for detecting FLAIR WMH change longitudinally after stroke. Methods: Ischemic stroke patients enrolled in an observational cohort study underwent serial MRI acutely, at 1 month and every 3 months thereafter. Patients with 6 to 12-month follow-up were included. WMH lesion volumes were automatically segmented on serial FLAIR MRI. In the hemisphere contralateral to the infarct after 30 days, we compared longitudinal change in WMH volume to a VW regression of FLAIR signal change over time. Serial FLAIR images were nonlinearly aligned using a T1 intermediate. FLAIR intensities were normalized, corrected for local expansion/contraction, and smoothed. Within voxels identified as lesion at any timepoint, significant WMH change was determined relative to the normal appearing WM and converted to percent of total intracranial volume (TICV). Results: 22 stroke patients (mean age 73, NIHSS 4, infarct volume 10.3cc) had on average 3 follow-ups over 343 days. In the non-stroke hemisphere, volumetry revealed lesion growth in 12 participants (55%) while VW analysis revealed FLAIR signal increase in 15 (68%). For volumetry, mean lesion increase was 0.01 ± 0.20% of TICV per year while VW FLAIR intensity increased in mean 0.11 ± 0.23% of total intracranial voxels. Volumetric growth was associated with the volume of VW signal increase (Std Beta = 0.51, p = 0.02). In the stroke hemisphere, WMH expansion of peri-infarct tissue, consistent with Wallerian degeneration, was present in 10 participants (45%). WMH regression, consistent with atrophy or infarct cavitation, was present in 14 participants (64%). Stroke hemisphere WMH expansion was associated with higher NIHSS (Std B = 0.58, p = 0.009) while regression was associated with larger infarct volumes (Std B = 0.80, p < 0.001). Conclusions: In early chronic stroke, VW longitudinal WMH analysis detects progressive WM degradation with regional variation and may be more sensitive than gross volumetry.

Assessment of higher-order singular value decomposition denoising methods on dynamic hyperpolarized [1-13C]pyruvate MRI data from patients with glioma

BackgroundReal-time metabolic conversion of intravenously-injected hyperpolarized [1-13C]pyruvate to [1-13C]lactate and [13C]bicarbonate in the brain can be measured using dynamic hyperpolarized carbon-13 (HP-13C) MRI. However, voxel-wise evaluation of metabolism in patients with glioma is challenged by the limited signal-to-noise ratio (SNR) of downstream 13C metabolites, especially within lesions. The purpose of this study was to evaluate the ability of higher-order singular value decomposition (HOSVD) denoising methods to enhance dynamic HP [1-13C]pyruvate MRI data acquired from patients with glioma. MethodsDynamic HP-13C MRI were acquired from 14 patients with glioma. The effects of two HOSVD denoising techniques, tensor rank truncation-image enhancement (TRI) and global–local HOSVD (GL-HOSVD), on the SNR and kinetic modeling were analyzed in [1-13C]lactate data with simulated noise that matched the levels of [13C]bicarbonate signals. Both methods were then evaluated in patient data based on their ability to improve [1-13C]pyruvate, [1-13C]lactate and [13C]bicarbonate SNR. The effects of denoising on voxel-wise kinetic modeling of kPL and kPB was also evaluated. The number of voxels with reliable kinetic modeling of pyruvate-to-lactate (kPL) and pyruvate-to-bicarbonate (kPB) conversion rates within regions of interest (ROIs) before and after denoising was then compared. ResultsBoth denoising methods improved metabolite SNR and regional signal coverage. In patient data, the average increase in peak dynamic metabolite SNR was 2-fold using TRI and 4–5 folds using GL-HOSVD denoising compared to acquired data. Denoising reduced kPL modeling errors from a native average of 23% to 16% (TRI) and 15% (GL-HOSVD); and kPB error from 42% to 34% (TRI) and 37% (GL-HOSVD) (values were averaged voxelwise over all datasets). In contrast-enhancing lesions, the average number of voxels demonstrating within-tolerance kPL modeling error relative to the total voxels increased from 48% in the original data to 84% (TRI) and 90% (GL-HOSVD), while the number of voxels showing within-tolerance kPB modeling error increased from 0% to 15% (TRI) and 8% (GL-HOSVD). ConclusionPost-processing denoising methods significantly improved the SNR of dynamic HP-13C imaging data, resulting in a greater number of voxels satisfying minimum SNR criteria and maximum kinetic modeling errors in tumor lesions. This enhancement can aid in the voxel-wise analysis of HP-13C data and thereby improve monitoring of metabolic changes in patients with glioma following treatment.

Advanced diffusion‐MRI models reveals sex‐specific differences in white matter organization in patients with mild cognitive impairment

AbstractBackgroundRecent literature suggests that free‐water corrected (fiso) diffusion MRI (dMRI) measures provide higher sensitivity as compared to conventional single‐tensor (ST) dMRI measures to understand white‐matter (WM) changes due to neurodegeneration. However, fiso‐corrected dMRI measures are still biased at crossing‐fiber regions (CFR) of WM which are shown to be around 90% of the total WM voxels with conventional dMRI data acquisition. In this study, we systematically compared dMRI measures from ST‐dMRI, fiso‐corrected ST‐dMRI measures, and measures from advanced dMRI models such as Diffusion Kurtosis Imaging (DKI) and Neurite Orientation Distribution and Density Imaging (NODDI) which are sensitive to CFRs. We hypothesize that these advanced dMRI models will be sensitive in understanding sex‐specific WM organizational changes at regional level in patients with Alzheimer’s disease (AD)‐related mild cognitive impairment (MCI).Method40 healthy controls [HC; 21 Females (F), 19 Males (M)] and 38 MCI participants [13F, 25 M] were recruited at our center. MCI diagnoses were based on expert consensus between a neurologist (AR) and a neuropsychologist (JC) using results of a neuropsychological test battery and the Clinical Dementia Rating (CDR) interview. Isotropic 1.5mm3 multi‐shell dMRI (b‐values=500s/mm2, 1000s/mm2, and 2500s/mm2) each with identical 71 diffusion‐encoding directions was acquired on a 3T Siemens Skyra scanner for every participant using the CMRR sequence. 8 uniformly interspersed non‐diffusion‐weighted (b0) images were also acquired in the same run. ST, fiso‐corrected, DKI, and NODDI dMRI measures were estimated for every brain voxel in‐house using the standard analytic tools. WM was skeletonized using FSL tools and statistical comparisons were done non‐parametrically using PALM tools. GLM was utilized to test for sex‐specific group effect. Significance was established at family‐wise error correction (FWE) pcorr<0.05.ResultThalamocortical tracts of males in both HC and MCI showed a significantly higher axonal kurtosis (AK) and orientation density (ODI) derived from DKI and NODDI when compared to within‐group females. DKI also revealed presence of higher axonal water fraction (AWF) along the entire WM‐tracts between HC‐female and HC‐male, and lower mean kurtosis (MK) along the thalamic‐cortical tracts in HC‐females as compared to MCI‐females.ConclusionAdvanced dMRI‐models are sensitive to understand sex‐specific WM disorganization due to healthy ageing and MCI.

CNTM-05. Left hemisphere gliomas induce a plastic bi-hemispheric language network further characterized by lobe specific glioma data

Abstract Primary brain tumors are among the most burdensome diagnoses patients can receive as they often carry with them externally obvious and significant detriments to motor and speech. The stroke model is severely limited by the inherent nature of the insult to the brain: binary, relatively instantaneous and defined by vascular boundaries. We instead have chosen to study glioma-induced neuroplasticity in patients with gliomas as presentation is over a significantly longer time course with a gradient of insult to language activation areas instead of immediate ablation. Chart review was conducted on 545 patients who completed fMRI imaging from 2013–2019 while completing 1+ of the following language tasks: antonyms, reading comprehension, rhyming, word generation and picture naming. A total of 117 patients’ fMRI scans were included in the final analysis which entailed both a cluster-based analysis in FSL and a 34 gyral, mask-based analysis using FEATquery. Right hemisphere gliomas (RHG) were first established as a reliable control group by averaging anatomically significant voxels across all five language tasks (LT) yielding a core conserved network. Left hemisphere gliomas (LHG) were then directly compared to the RHG language tasks. We found LHG induced global reorganization of the conserved language network with little evidence for direct homologous recruitment of functional structures. Instead, a generalized right hemisphere recruitment is observed with 87% of non-zero masks shifting their laterality index to the right hemisphere. Furthermore, in each of the five LT, the LHG activates fewer total suprathreshold voxels in both the mask and cluster based analyses while having a higher peak intensity within the activated clusters. A preliminary analysis of frontal LHG compared to temporal LHG reveals increased contralateral recruitment in the frontal subgroup with more direct homologous recruitment. This nuanced understanding of existing mechanisms for neuroplasticity can aid in our future intentional manipulation for therapeutic benefit.

Data Quality Assessment for Super-Resolution Fetal Brain MR Imaging: A Retrospective 1.5 T Study.

Super-resolution is a promising technique to create isotropic image volumes from stacks of two-dimensional (2D) motion-corrupted images in fetal magnetic resonance imaging (MRI). To determine an acquisition quality metric and correlate that metric with radiologist perception of three-dimensional (3D) image quality. Retrospective. Eighty-seven patients, mean gestational age 29 ± 6 weeks. 1.5 T/2D fast spin-echo. Four radiologists (L.G., D.M.E.B., P.C., and J.V.; 31, 21, 7, and 7 years' experience, respectively) graded reconstructions on a 0 to 4 scale for overall appearance and visibility of specific anatomy. During reconstruction, slices were labeled as inliers based on correlation between a simulated vs. actual acquisition. The fraction of brain voxels in inlier slicers vs. total brain voxels was measured for each acquisition. Paired sample t test, Pearson's correlation, intra-class correlation. The average brain mask inlier fraction for all acquisitions was 0.8. There was a statistically significant correlation (0.71) between overall reconstruction appearance and number of acquisitions with inlier fraction above 0.73. There was low correlation (0.21, P= 0.05) between the number of acquisitions used in the reconstruction and overall score when no data quality measure was considered. Similar results were found for ratings of specific anatomy. Statistically significant differences in overall perception of image quality were found when using three vs. four, four vs. five, and three vs. five high-quality acquisitions in the reconstruction. Five high-quality acquisitions were sufficient to yield an average radiologist rating of 3.59 out of 4.0 for overall image quality. Reconstruction quality can be reliably predicted using the brain mask inlier fraction. Real-time super-resolution protocols could exploit this to terminate acquisition when enough high-quality acquisitions have been collected. To achieve consistent 3D image quality it may be necessary to acquire more than five scans to compensate for severely motion-corrupted acquisitions. 3 TECHNICAL EFFICACY: 1.

Ground-truth “resting-state” signal provides data-driven estimation and correction for scanner distortion of fMRI time-series dynamics

The fMRI community has made great strides in decoupling neuronal activity from other physiologically induced T2* changes, using sensors that provide a ground-truth with respect to cardiac, respiratory, and head movement dynamics. However, blood oxygenation level-dependent (BOLD) time-series dynamics are also confounded by scanner artifacts, in complex ways that can vary not only between scanners but even, for the same scanner, between sessions. Unfortunately, the lack of an equivalent ground truth for BOLD time-series has thus far stymied the development of reliable methods for identification and removal of scanner-induced noise, a problem that we have previously shown to severely impact detection sensitivity of resting-state brain networks. To address this problem, we first designed and built a phantom capable of providing dynamic signals equivalent to that of the resting-state brain. Using the dynamic phantom, we then compared the ground-truth time-series with its measured fMRI data. Using these, we introduce data-quality metrics: Standardized Signal-to-Noise Ratio (ST-SNR) and Dynamic Fidelity that, unlike currently used measures such as temporal SNR (tSNR), can be directly compared across scanners. Dynamic phantom data acquired from four “best-case” scenarios: high-performance scanners with MR-physicist-optimized acquisition protocols, still showed scanner instability/multiplicative noise contributions of about 6–18% of the total noise. We further measured strong non-linearity in the fMRI response for all scanners, ranging between 8–19% of total voxels. To correct scanner distortion of fMRI time-series dynamics at a single-subject level, we trained a convolutional neural network (CNN) on paired sets of measured vs. ground-truth data. The CNN learned the unique features of each session's noise, providing a customized temporal filter. Tests on dynamic phantom time-series showed a 4- to 7-fold increase in ST-SNR and about 40–70% increase in Dynamic Fidelity after denoising, with CNN denoising outperforming both the temporal bandpass filtering and denoising using Marchenko-Pastur principal component analysis. Critically, we observed that the CNN temporal denoising pushes ST-SNR to a regime where signal power is higher than that of noise (ST-SNR > 1). Denoising human-data with ground-truth-trained CNN, in turn, showed markedly increased detection sensitivity of resting-state networks. These were visible even at the level of the single-subject, as required for clinical applications of fMRI.

A Potential Biomarker from Diffusion Weighted Imaging and Parametric Response Map Analysis for Treatment Response Prediction in Nasopharyngeal Cancer

Diffusion-weighted imaging (DWI) is an MRI technique which provides functional information of tissue by detecting microscopic motion of water molecules. The change of apparent diffusion coefficient (ADC) derived from DWI was used as an imaging biomarker for treatment response prediction in cancers [1]. However, it was based on whole-tumor analysis which did not reflect heterogeneity within the tumor. To overcome this limitation, a new method called parametric response map (PRM) analysis was proposed to evaluate response by quantifying voxel-wise changes in ADC [2]. Here we investigated the use of PRM analysis on ADC from DWI as an imaging biomarker for treatment response prediction in nasopharyngeal cancer (NPC) patients. We collected thirteen patient datasets including ten complete response (CR) patients and three partial response (PR) patients at King Chulalongkorn Memorial Hospital where one patient dataset consisted of DWI and ADC data acquired before (i.e. pre-treatment) and at five weeks after (i.e. mid-treatment) initiation of chemoradiation therapy. For each dataset, we compared pre-treatment ADC image with co-registered mid-treatment ADC image, and calculated the percentage of voxels with increased ADC values with respect to total voxels within the tumor ROI, defined as PRM+. To validate the feasibility of the PRM biomarker, we computed the mean and standard deviation (SD) of percentage change in tumor volume (%AVol ) and in ADC (%AADC) and PRM+ across CR and PR patients, where tumor response was from 6-month follow-up data using RECIST1.1 guideline. The results showed that %AVol as well as %AADC between both groups was not significantly different. In contrast, PRM+ was significantly different between both groups (p < 0.05, 82.7±7.8% in CR vs 66.7±6.5% in PR). Our results implied that the proposed PRM+ biomarker could be a potential biomarker for early treatment response prediction in NPC patients.

3D multifractal characterization of computed tomography images of soils under different tillage management: Linking multifractal parameters to physical properties

Multifractal analysis of pore images obtained from X-ray computed tomography (CT) was used to characterize the scaling properties of macropores in soils with different managements and their correspondence with macroscopic physical properties related with the soil functions.We used CT images of twenty undisturbed soil columns to examine the multifractal properties of the pores identified by X-ray computed tomography (CT-Porosity). Multifractal spectra successfully describe the scaling of the pore network in all soil columns. The dimensions and scaling parameters of these spectra correlate with macroscopic magnitudes, namely, CT-Porosity, surface area of the pore walls, tortuosity, and bulk density. We also found strong correlations between the singularity spectra and the topological descriptors of the pore network skeleton: total slab voxels, number of branches per path, number of endpoints and sum of branch length, among others. These correlations show that the complexity of the CT-Porosity can be related quantitatively with physical properties, the organization of the pore skeleton and solute transport.

Local complexity metrics to quantify the effect of anatomical noise on detectability of lung nodules in chest CT imaging.

The purpose of this study is to (1)develop metrics to characterize the regional anatomical complexity of the lungs, and (2)relate these metrics with lung nodule detection in chest CT. A free-scrolling reader-study with virtually inserted nodules (13 radiologists × 157 total nodules = 2041 responses) is used to characterize human detection performance. Metrics of complexity based on the local density and orientation of distracting vasculature are developed for two-dimensional (2-D) and three-dimensional (3-D) considerations of the image volume. Assessed characteristics included the distribution of 2-D/3-D vessel structures of differing orientation (dubbed "2-D/3-D and dot-like/line-like distractor indices"), contiguity of inserted nodules with local vasculature, mean local gray-level surrounding each nodule, the proportion of lung voxels to total voxels in each section, and 3-D distance of each nodule from the trachea bifurcation. A generalized linear mixed-effects statistical model is used to determine the influence of each these metrics on nodule detectability. In order of decreasing effect size: 3-D line-like distractor index, 2-D line-like distractor index, 2-D dot-like distractor index, local mean gray-level, contiguity with 2-D dots, lung area, and contiguity with 3-D lines all significantly affect detectability ( ). These data demonstrate that local lung complexity degrades detection of lung nodules.

SU‐E‐T‐558: Monte Carlo Photon Transport Simulations On GPU with Quadric Geometry

Purpose:Monte Carlo simulation on GPU has experienced rapid advancements over the past a few years and tremendous accelerations have been achieved. Yet existing packages were developed only in voxelized geometry. In some applications, e.g. radioactive seed modeling, simulations in more complicated geometry are needed. This abstract reports our initial efforts towards developing a quadric geometry module aiming at expanding the application scope of GPU‐based MC simulations.Methods:We defined the simulation geometry consisting of a number of homogeneous bodies, each specified by its material composition and limiting surfaces characterized by quadric functions. A tree data structure was utilized to define geometric relationship between different bodies. We modified our GPU‐based photon MC transport package to incorporate this geometry. Specifically, geometry parameters were loaded into GPU's shared memory for fast access. Geometry functions were rewritten to enable the identification of the body that contains the current particle location via a fast searching algorithm based on the tree data structure.Results:We tested our package in an example problem of HDR‐brachytherapy dose calculation for shielded cylinder. The dose under the quadric geometry and that under the voxelized geometry agreed in 94.2% of total voxels within 20% isodose line based on a statistical t‐test (95% confidence level), where the reference dose was defined to be the one at 0.5cm away from the cylinder surface. It took 243sec to transport 100million source photons under this quadric geometry on an NVidia Titan GPU card. Compared with simulation time of 99.6sec in the voxelized geometry, including quadric geometry reduced efficiency due to the complicated geometry‐related computations.Conclusion:Our GPU‐based MC package has been extended to support photon transport simulation in quadric geometry. Satisfactory accuracy was observed with a reduced efficiency. Developments for charged particle transport in this geometry are currently in progress.

Pre-Radiation Therapy Fluorine 18 Fluorodeoxyglucose PET Helps Identify Patients with Esophageal Cancer at High Risk for Radiation Pneumonitis.

To examine the association between pre-radiation therapy (RT) fluorine 18 fluorodeoxyglucose (FDG) uptake and post-RT symptomatic radiation pneumonitis (RP). In accordance with the retrospective study protocol approved by the institutional review board, 228 esophageal cancer patients who underwent FDG PET/CT before chemotherapy and RT were examined. RP symptoms were evaluated by using the Common Terminology Criteria for Adverse Events, version 4.0, from the consensus of five clinicians. By using the cumulative distribution of standardized uptake values (SUVs) within the lungs, those values greater than 80%-95% of the total lung voxels were determined for each patient. The effect of pre-chemotherapy and RT FDG uptake, dose, and patient or treatment characteristics on RP toxicity was studied by using logistic regression. The study subjects were treated with three-dimensional conformal RT (n = 36), intensity-modulated RT (n = 135), or proton therapy (n = 57). Logistic regression analysis demonstrated elevated FDG uptake at pre-chemotherapy and RT was related to expression of RP symptoms. Study subjects with elevated 95% percentile of the SUV (SUV95) were more likely to develop symptomatic RP (P < .000012); each 0.1 unit increase in SUV95 was associated with a 1.36-fold increase in the odds of symptomatic RP. Receiver operating characteristic (ROC) curve analysis resulted in area under the ROC curve of 0.676 (95% confidence interval: 0.58, 0.77), sensitivity of 60%, and specificity of 71% at the 1.17 SUV95 threshold. CT imaging and dosimetric parameters were found to be poor predictors of RP symptoms. The SUV95, a biomarker of pretreatment pulmonary metabolic activity, was shown to be prognostic of symptomatic RP. Elevation in this pretreatment biomarker identifies patients at high risk for posttreatment symptomatic RP.

Tropospheric wet refractivity tomography based on the BeiDou satellite system

This paper presents a novel approach for assessing the precision of the wet refractivity field using BDS (BeiDou navigation satellite system) simulations only, GPS, and BDS+GPS for the Shenzhen and Hongkong GNSS network. The simulations are carried out by adding artificial noise to a real observation dataset. Instead of using the d and s parameters computed from slant wet delay, as in previous studies, we employ the Bias and RMS parameters, computed from the tomography results of total voxels, in order to obtain a more direct and comprehensive evaluation of the precision of the refractivity field determination. The results show that: (1) the precision of tropospheric wet refractivity estimated using BDS alone (only 9 satellites used) is basically comparable to that of GPS; (2) BDS+GPS (as of current operation) may not be able to significantly improve the data’s spatial density for the application of refractivity tomography; and (3) any slight increase in the precision of refractivity tomography, particularly in the lower atmosphere, bears great significance for any applications dependent on the Chinese operational meteorological service.

The Effect of Small Tumor Volumes on Studies of Intratumoral Heterogeneity of Tracer Uptake

The number of studies in the literature involving quantification of the metabolic heterogeneity seen in (18)F-FDG PET images has increased sharply over recent years. We hypothesized that inclusion of very small regions of interest as unique data points will have deleterious effects on these studies. Using a combination of probability theory and clinical (18)F-FDG PET data, we numerically calculated the curve describing the probability a given tumor volume is large enough to adequately sample the underlying tumor biology assayed via a PET/CT scanner at a planar resolution of 4 mm and transaxial resolution of 4 mm (64 mm(3) voxel size). We then used a computer simulation to isolate the effects of tumor volume on the image local entropy. We computed the underlying global intensity distribution for 70 cervical cancer tumors ranging from 4 to 248 cm(3)), which were ensemble-averaged over the same intensity scale. From this distribution, we determined that about 700 total voxels (45 cm(3)) are required to give 95% certainty that the global intensity distribution has been sufficiently sampled for common statistical comparisons of individual tumor intensity distributions to be made canonically. We demonstrated that one previously suggested measure of heterogeneity is dependent on tumor volume and that measurement of heterogeneity is about 5 times more sensitive to volume changes for volumes below the proposed minimum than for those above it. Inclusion of tumor volumes below 45 cm(3) can profoundly bias comparisons of intratumoral uptake heterogeneity metrics derived from data from the current generation of whole-body (18)F-FDG PET scanners.