Parcellation Method Research Articles

CartiMorph: A framework for automated knee articular cartilage morphometrics

We introduce CartiMorph, a framework for automated knee articular cartilage morphometrics. It takes an image as input and generates quantitative metrics for cartilage subregions, including the percentage of full-thickness cartilage loss (FCL), mean thickness, surface area, and volume. CartiMorph leverages the power of deep learning models for hierarchical image feature representation. Deep learning models were trained and validated for tissue segmentation, template construction, and template-to-image registration. We established methods for surface-normal-based cartilage thickness mapping, FCL estimation, and rule-based cartilage parcellation. Our cartilage thickness map showed less error in thin and peripheral regions. We evaluated the effectiveness of the adopted segmentation model by comparing the quantitative metrics obtained from model segmentation and those from manual segmentation. The root-mean-squared deviation of the FCL measurements was less than 8%, and strong correlations were observed for the mean thickness (Pearson’s correlation coefficient ρ∈[0.82,0.97]), surface area (ρ∈[0.82,0.98]) and volume (ρ∈[0.89,0.98]) measurements. We compared our FCL measurements with those from a previous study and found that our measurements deviated less from the ground truths. We observed superior performance of the proposed rule-based cartilage parcellation method compared with the atlas-based approach. CartiMorph has the potential to promote imaging biomarkers discovery for knee osteoarthritis.

The effects of aerosol concentration on the evolution, transport, and deposition of hygroscopic droplets in the highly idealized MT model: A numerical study

Inhalation therapy is widely used to treat various respiratory diseases. Nebulizers have been widely adopted due to their ease of operation, high drug delivery efficiency, and low residual volume. Traditional studies on drug delivery in the respiratory tract typically use a one-way coupled method to track drug transport and deposition for non-hygroscopic drug particles. However, neglecting the concentration of nebulized hygroscopic droplet aerosols can lead to significant discrepancies in the deposition efficiency (DE) of nebulized drugs in the airways. To investigate the effect of different nebulization rates on the evolution, transport, and deposition of droplets, a highly idealized mouth-throat (MT) model was selected as a testbed for the in silico trials of drug delivery to the respiratory tract. The interaction between continuous and discrete phases was modeled using a two-way coupled model of highly concentrated droplet aerosol-water vapor interaction, innovated by the particle parcel method. The effects of different coupling methods and boundary conditions on the average droplet diameter and droplet deposition distribution pattern at the outlet of the MT model were compared using a computational fluid dynamics (CFD) method. Simulation results indicate that neglecting the effect of water vapor evaporated from the droplets on the relative humidity (RH) in the oral cavity could lead to a 2.2-fold prediction error of the average droplet diameter at the outlet of the MT model for a conventional nebulization rate of 0.386 g/min. As the nebulization rate increases, the difference in the predicted average droplet diameter at the outlet of the MT model between the one-way coupled method and the two-way coupled method also increases. Increasing the nebulization rate during inhalation therapy could lead to incomplete evaporation of drug droplets, resulting in deposition of larger diameter drug droplets in the oral cavity. In contrast, a lower nebulization rate can reduce the droplet diameter at the outlet of the MT model, allowing it to enter deeper airways.

Self-esteem and cortical thickness correlate with aggression in healthy children: A surface‐based analysis

Aggressive behavior can have serious physical, psychological, and social consequences. However, little is known about the personality and neurological antecedents underlying aggressive behavior in children. The objective of this study was to investigate the relationship between self-esteem, aggression, and brain structure (i.e., cortical thickness and surface area) in a population of healthy children (N = 78; 9–12 years; mean age: 9.95 ± 0.90 years). The results revealed that self-esteem showed a negative association with aggression and significantly predicted aggressive behavior. No gender differences were found in aggression and its neural correlates. We performed the cortical parcellation method to further explore the neural foundations underlying the association of self-esteem with aggression. Children with higher aggression had increased cortical thickness in four clusters after multiple comparison correction: right medial orbitofrontal cortex, right lateral orbitofrontal cortex, right superior frontal gyrus, and left insula. In a mediation analysis, cortical thickness in the right medial orbitofrontal cortex contributed to the effect of self-esteem on aggression. These findings extend our understanding of morphological correlates of aggression in children, suggesting that an increased cortical thickness in childhood is a potential mechanism linking low self-esteem to aggression.

Mapping corpus callosum surface reduction in fetal alcohol spectrum disorders with sulci and connectivity-based parcellation.

Fetal alcohol spectrum disorders (FASD) range from fetal alcohol syndrome (FAS) to non-syndromic non-specific forms (NS-FASD) that are still underdiagnosed and could benefit from new neuroanatomical markers. The main neuroanatomical manifestation of prenatal alcohol exposure on developmental toxicity is the reduction in brain size, but repeated imaging observations have long driven the attention on the corpus callosum (CC), without being all convergent. Our study proposed a new segmentation of the CC that relies on both a sulci-based cortical segmentation and the "hemispherotopic" organization of the transcallosal fibers. We collected a monocentric series of 37 subjects with FAS, 28 with NS-FASD, and 38 with typical development (6 to 25 years old) using brain MRI (1.5T). Associating T1- and diffusion-weighted imaging, we projected a sulci-based cortical segmentation of the hemispheres on the midsagittal section of the CC, resulting in seven homologous anterior-posterior parcels (frontopolar, anterior and posterior prefrontal, precentral, postcentral, parietal, and occipital). We measured the effect of FASD on the area of callosal and cortical parcels by considering age, sex, and brain size as linear covariates. The surface proportion of the corresponding cortical parcel was introduced as an additional covariate. We performed a normative analysis to identify subjects with an abnormally small parcel. All callosal and cortical parcels were smaller in the FASD group compared with controls. When accounting for age, sex, and brain size, only the postcentral (η2 = 6.5%, pFDR = 0.032) callosal parcel and % of the cortical parcel (η2 = 8.9%, pFDR = 0.007) were still smaller. Adding the surface proportion (%) of the corresponding cortical parcel to the model, only the occipital parcel was persistently reduced in the FASD group (η2 = 5.7%, pFDR = 0.014). In the normative analysis, we found an excess of subjects with FASD with abnormally small precentral and postcentral (peri-isthmic) and posterior-splenial parcels (pFDR < 0.05). The objective sulcal and connectivity-based method of CC parcellation proved to be useful not only in confirming posterior-splenial damage in FASD but also in the narrowing of the peri-isthmic region strongly associated with a specific size reduction in the corresponding postcentral cortical region (postcentral gyrus). The normative analysis showed that this type of callosal segmentation could provide a clinically relevant neuroanatomical endophenotype, even in NS-FASD.

Hippocampal anterior- posterior shift in childhood and adolescence.

Hippocampal-cortical networks play an important role in neurocognitive development. Applying the method of Connectivity-Based Parcellation (CBP) on hippocampal-cortical structural covariance (SC) networks computed from T1-weighted magnetic resonance images, we examined how the hippocampus differentiates into subregions during childhood and adolescence (N=1105, 6-18 years). In late childhood, the hippocampus mainly differentiated along the anterior-posterior axis similar to previous reported functional differentiation patterns of the hippocampus. In contrast, in adolescence a differentiation along the medial-lateral axis was evident, reminiscent of the cytoarchitectonic division into cornu ammonis and subiculum. Further meta-analytical characterization of hippocampal subregions in terms of related structural co-maturation networks, behavioural and gene profiling suggested that the hippocampal head is related to higher order functions (e.g. language, theory of mind, autobiographical memory) in late childhood morphologically co-varying with almost the whole brain. In early adolescence but not in childhood, posterior subicular SC networks were associated with action-oriented and reward systems. The findings point to late childhood as an important developmental period for hippocampal head morphology and to early adolescence as a crucial period for hippocampal integration into action- and reward-oriented cognition. The latter may constitute a developmental feature that conveys increased propensity for addictive disorders.

Superficial white matter analysis: An efficient point-cloud-based deep learning framework with supervised contrastive learning for consistent tractography parcellation across populations and dMRI acquisitions.

Diffusion MRI tractography is an advanced imaging technique that enables in vivo mapping of the brain's white matter connections. White matter parcellation classifies tractography streamlines into clusters or anatomically meaningful tracts. It enables quantification and visualization of whole-brain tractography. Currently, most parcellation methods focus on the deep white matter (DWM), whereas fewer methods address the superficial white matter (SWM) due to its complexity. We propose a novel two-stage deep-learning-based framework, Superficial White Matter Analysis (SupWMA), that performs an efficient and consistent parcellation of 198 SWM clusters from whole-brain tractography. A point-cloud-based network is adapted to our SWM parcellation task, and supervised contrastive learning enables more discriminative representations between plausible streamlines and outliers for SWM. We train our model on a large-scale tractography dataset including streamline samples from labeled long- and medium-range (over 40mm) SWM clusters and anatomically implausible streamline samples, and we perform testing on six independently acquired datasets of different ages and health conditions (including neonates and patients with space-occupying brain tumors). Compared to several state-of-the-art methods, SupWMA obtains highly consistent and accurate SWM parcellation results on all datasets, showing good generalization across the lifespan in health and disease. In addition, the computational speed of SupWMA is much faster than other methods.

Structural MRI predicts clinical progression in presymptomatic genetic frontotemporal dementia: findings from the GENetic Frontotemporal dementia Initiative cohort.

Biomarkers that can predict disease progression in individuals with genetic frontotemporal dementia are urgently needed. We aimed to identify whether baseline MRI-based grey and white matter abnormalities are associated with different clinical progression profiles in presymptomatic mutation carriers in the GENetic Frontotemporal dementia Initiative. Three hundred eighty-seven mutation carriers were included (160 GRN, 160 C9orf72, 67 MAPT), together with 240 non-carrier cognitively normal controls. Cortical and subcortical grey matter volumes were generated using automated parcellation methods on volumetric 3T T1-weighted MRI scans, while white matter characteristics were estimated using diffusion tensor imaging. Mutation carriers were divided into two disease stages based on their global CDR®+NACC-FTLD score: presymptomatic (0 or 0.5) and fully symptomatic (1 or greater). The w-scores in each grey matter volumes and white matter diffusion measures were computed to quantify the degree of abnormality compared to controls for each presymptomatic carrier, adjusting for their age, sex, total intracranial volume, and scanner type. Presymptomatic carriers were classified as 'normal' or 'abnormal' based on whether their grey matter volume and white matter diffusion measure w-scores were above or below the cut point corresponding to the 10th percentile of the controls. We then compared the change in disease severity between baseline and one year later in both the 'normal' and 'abnormal' groups within each genetic subtype, as measured by the CDR®+NACC-FTLD sum-of-boxes score and revised Cambridge Behavioural Inventory total score. Overall, presymptomatic carriers with normal regional w-scores at baseline did not progress clinically as much as those with abnormal regional w-scores. Having abnormal grey or white matter measures at baseline was associated with a statistically significant increase in the CDR®+NACC-FTLD of up to 4 points in C9orf72 expansion carriers, and 5 points in the GRN group as well as a statistically significant increase in the revised Cambridge Behavioural Inventory of up to 11 points in MAPT, 10 points in GRN, and 8 points in C9orf72 mutation carriers. Baseline regional brain abnormalities on MRI in presymptomatic mutation carriers are associated with different profiles of clinical progression over time. These results may be helpful to inform stratification of participants in future trials.

Enhanced brain parcellation via abnormality inpainting for neuroimage-based consciousness evaluation of hydrocephalus patients by lumbar drainage

Brain network analysis based on structural and functional magnetic resonance imaging (MRI) is considered as an effective method for consciousness evaluation of hydrocephalus patients, which can also be applied to facilitate the ameliorative effect of lumbar cerebrospinal fluid drainage (LCFD). Automatic brain parcellation is a prerequisite for brain network construction. However, hydrocephalus images usually have large deformations and lesion erosions, which becomes challenging for ensuring effective brain parcellation works. In this paper, we develop a novel and robust method for segmenting brain regions of hydrocephalus images. Our main contribution is to design an innovative inpainting method that can amend the large deformations and lesion erosions in hydrocephalus images, and synthesize the normal brain version without injury. The synthesized images can effectively support brain parcellation tasks and lay the foundation for the subsequent brain network construction work. Specifically, the novelty of the inpainting method is that it can utilize the symmetric properties of the brain structure to ensure the quality of the synthesized results. Experiments show that the proposed brain abnormality inpainting method can effectively aid the brain network construction, and improve the CRS-R score estimation which represents the patient’s consciousness states. Furthermore, the brain network analysis based on our enhanced brain parcellation method has demonstrated potential imaging biomarkers for better interpreting and understanding the recovery of consciousness in patients with secondary hydrocephalus.

Influence of accessory sulci of the frontoparietal operculum on gray matter quantification.

Introduction: The perisylvian region is the cortical core of language and speech. Several accessory sulci have been described in this area, whose presence could modify the results of the automatic quantification of gray matter by popularly used software. This study aimed to assess the expression of accessory sulci in the frontoparietal operculum (FPO) and to evaluate their influence on the gray matter volume estimated by an automatic parcellation of cortical gyri and sulci. Methods: Brain MRI scans of 100 healthy adult volunteers were visually analyzed. The existence of the triangular and diagonal sulci, and the number of accessory sulci in the frontoparietal operculum, were assessed on T1 images. Also, the gray matter volume of gyri and sulci was quantified by an automatized parcellation method. Interhemispheric differences in accessory sulci were evaluated with Chi-square and Wilcoxon paired tests. The effects of the hemisphere, sex, age, total intracranial volume, and accessory sulci on morphometric variables were assessed by linear models. Results: These sulci were found in more than half of the subjects, mostly in the left hemisphere, and showed a significant effect on the gray matter content of the FPO. In particular, the volume of the inferior frontal sulcus, pars opercularis of the inferior frontal gyrus, horizontal ramus of the lateral sulcus, angular gyrus, and postcentral gyrus showed a significant influence on the presence of accessory sulci. Discussion: The prevalence of tertiary sulci in the FPO is high, although their meaning is not yet known. Therefore, they should be considered to reduce the risk of misclassifications of normal variation.

DeepParcellation: A novel deep learning method for robust brain magnetic resonance imaging parcellation in older East Asians.

Accurate parcellation of cortical regions is crucial for distinguishing morphometric changes in aged brains, particularly in degenerative brain diseases. Normal aging and neurodegeneration precipitate brain structural changes, leading to distinct tissue contrast and shape in people aged >60 years. Manual parcellation by trained radiologists can yield a highly accurate outline of the brain; however, analyzing large datasets is laborious and expensive. Alternatively, newly-developed computational models can quickly and accurately conduct brain parcellation, although thus far only for the brains of Caucasian individuals. To develop a computational model for the brain parcellation of older East Asians, we trained magnetic resonance images of dimensions 256 × 256 × 256 on 5,035 brains of older East Asians (Gwangju Alzheimer's and Related Dementia) and 2,535 brains of Caucasians. The novel N-way strategy combining three memory reduction techniques inception blocks, dilated convolutions, and attention gates was adopted for our model to overcome the intrinsic memory requirement problem. Our method proved to be compatible with the commonly used parcellation model for Caucasians and showed higher similarity and robust reliability in older aged and East Asian groups. In addition, several brain regions showing the superiority of the parcellation suggest that DeepParcellation has a great potential for applications in neurodegenerative diseases such as Alzheimer's disease.

HOA2.0-ComPaRe: A next generation Harvard-Oxford Atlas comparative parcellation reasoning method for human and macaque individual brain parcellation and atlases of the cerebral cortex.

Comparative structural neuroanatomy is a cornerstone for understanding human brain structure and function. A parcellation framework that relates systematically to fundamental principles of histological organization is an essential step in generating structural comparisons between species. In the present investigation, we developed a comparative parcellation reasoning system (ComPaRe), which is a formal ontological system in human and non-human primate brains based on the cortical cytoarchitectonic mapping used for both species as detailed by Brodmann. ComPaRe provides a theoretical foundation for mapping neural systems in humans and other species using neuroimaging. Based on this approach, we revised the methodology of the original Harvard-Oxford Atlas (HOA) system of brain parcellation to produce a comparative framework for the human (hHOA) and the rhesus monkey (mHOA) brains, which we refer to as HOA2.0-ComPaRe. In addition, we used dedicated segmentation software in the publicly available 3D Slicer platform to parcellate an individual human and rhesus monkey brain. This method produces quantitative morphometric parcellations in the individual brains. Based on these parcellations we created a representative template and 3D brain atlas for the two species, each based on a single subject. Thus, HOA2.0-ComPaRe provides a theoretical foundation for mapping neural systems in humans and other species using neuroimaging, while also representing a significant revision of the original human and macaque monkey HOA parcellation schemas. The methodology and atlases presented here can be used in basic and clinical neuroimaging for morphometric (volumetric) analysis, further generation of atlases, as well as localization of function and structural lesions.

Data-Driven Clustering of Functional Signals Reveals Gradients in Processing Both within the Anterior Hippocampus and across Its Long Axis.

A particularly elusive puzzle concerning the hippocampus is how the structural differences along its long anteroposterior axis might beget meaningful functional differences, particularly in terms of the granularity of information processing. One measure posits to quantify this granularity by calculating the average statistical independence of the BOLD signal across neighboring voxels, or intervoxel similarity (IVS), and has shown the anterior hippocampus to process coarser-grained information than the posterior hippocampus. This measure, however, has yielded opposing results in studies of developmental and healthy aging samples, which also varied in fMRI acquisition parameters and hippocampal parcellation methods. To reconcile these findings, we measured IVS across two separate resting-state fMRI acquisitions and compared the results across many of the most widely used parcellation methods in a large young-adult sample of male and female humans (Acquisition 1, N = 233; Acquisition 2, N = 176). Finding conflicting results across acquisitions and parcellations, we reasoned that a data-driven approach to hippocampal parcellation is necessary. To this end, we implemented a group masked independent components analysis to identify functional subunits of the hippocampus, most notably separating the anterior hippocampus into separate anterior-medial, anterior-lateral, and posteroanterior-lateral components. Measuring IVS across these components revealed a decrease in IVS along the medial-lateral axis of the anterior hippocampus but an increase from anterior to posterior. We conclude that intervoxel similarity is deeply affected by parcellation and that grounding one's parcellation in a functionally informed approach might allow for a more complex and reliable characterization of the hippocampus.SIGNIFICANCE STATEMENT Processing information along hierarchical scales of granularity is critical for many of the feats of cognition considered most human. Recently, the changes in structure, cortical connectivity, and apparent functional properties across parcels of the hippocampal long axis have been hypothesized to underlie this hierarchical gradient in information processing. We show here, however, that the choice of parcellation method itself drastically affects one particular measure of granularity across the hippocampus and that a functionally informed approach to parcellation reveals gradients both within the anterior hippocampus and in nonlinear form across the long axis. These results point to the issue of parcellation as a critical one in the study of the hippocampus and reorient interpretation of existing results.

Modeling Mixing in Stratified Heterogeneous Media: The Role of Water Velocity Discretization in Phase Space Formulation

Modeling solute transport in heterogeneous porous media faces two challenges: scale dependence of dispersion and reproducing mixing separately from spreading. Both are crucial since real applications may require km scales whereas reactions, often controlled by mixing, may occur at the pore scale. Methods have been developed in response to these challenges, but none has satisfactorily characterized both processes. In this paper, we propose a formulation based on the Water Mixing Approach extended to account for velocity variability. Velocity is taken as an independent variable, so that concentration depends on time, space and velocity. Therefore, we term the formulation the Multi-Advective Water Mixing Approach. A new mixing term between velocity classes emerges in this formulation. We test it on Poiseuille’s stratified flow using the Water Parcel method. Results show high accuracy of the formulation in both dispersion and mixing. Moreover, the mixing process exhibits Markovianity in space even though it is modeled in time.

Automatic method for individual parcellation of manganese-enhanced magnetic resonance imaging of rat brain.

AimsTo construct an automatic method for individual parcellation of manganese-enhanced magnetic resonance imaging (MEMRI) of rat brain with high accuracy, which could preserve the inherent voxel intensity and Regions of interest (ROI) morphological characteristics simultaneously.Methods and resultsThe transformation relationship from standardized space to individual space was obtained by firstly normalizing individual image to the Paxinos space and then inversely transformed. On the other hand, all the regions defined in the atlas image were separated and resaved as binary mask images. Then, transforming the mask images into individual space via the inverse transformations and reslicing using the 4th B-spline interpolation algorithm. The boundary of these transformed regions was further refined by image erosion and expansion operator, and finally combined together to generate the individual parcellations. Moreover, two groups of MEMRI images were used for evaluation. We found that the individual parcellations were satisfied, and the inherent image intensity was preserved. The statistical significance of case-control comparisons was further optimized.ConclusionsWe have constructed a new automatic method for individual parcellation of rat brain MEMRI images, which could preserve the inherent voxel intensity and further be beneficial in case-control statistical analyses. This method could also be extended to other imaging modalities, even other experiments species. It would facilitate the accuracy and significance of ROI-based imaging analyses.

Understanding high polluted events in a tropical megacity with air quality observations and ERA5-reanalysis data

Already published studies found that 20% of the atmospheric pollution variations are linked to local meteorology. This relationship may be more important in places with dense populations such as Latin American cities since the topography and fast multiscale changes are part of the tropical climate. Even so, this possibility has not been addressed in previous studies. This research aims to characterize the relationship between tropical climate variables and PM2.5 levels during high pollution events. The relationship between wind field, Turbulent Kinetic Energy (TKE), radiation, temperature, relative humidity, boundary layer height, and atmospheric stability with PM2.5 concentration was investigated. Statistical correlations and the parcel method were used to analyze the relationship between vertical motions and PM2.5. Obtained results show that the stability, vertical velocity, and boundary layer height do not significantly affect pollution levels. We identified some signals that are strongly related to PM2.5 high concentrations: weaker than average horizontal wind speed and TKE throughout the day, easterly winds in the morning (associated with the transport of ashes from wildfires produced on the Eastern plains), combined with a higher than average radiation peak. These results lead to a better understanding of the PM2.5 variations, which can be applied for the improvement of air quality models and have the potential to be part of a novel policy to manage air quality risk.

Land Protection Partnership in the Montezuma Wetlands Complex, New York, USA

Conserving critical wildlife habitat at a regional scale can be challenging, especially when the region hosts a range of land uses, jurisdictions, and competing interests. Abundant opportunities exist for cooperation when vested conservation entities find common ground to use their unique strengths in a cooperative effort to protect and restore wetlands for wildlife and people. We present the Montezuma Wetlands Complex (MWC) Land Protection Partnership as a case study of regional conservation collaboration aimed at identifying areas in greatest need of wetland protection and restoration to support wetland wildlife and provide wildlife-based recreation. The MWC is among the most important wetland complexes in the Atlantic flyway of eastern North America for migratory birds because it provides critical migratory stopover habitat for millions of birds and regionally unique habitats for breeding birds and resident wildlife, including numerous endangered and threatened (E&amp;T) species. This case study demonstrates how state, federal, and nonprofit entities with differing goals and objectives can partner to protect and restore critical wetland habitat for wildlife. Partners optimized efforts by developing an online survey that included physical, land cover, biological, and people/use attributes which were ranked by each partner to determine common priorities and applied these into a spatial mapping, decision-support tool. Within attribute categories, land protection (physical), emergent marshes (land use), E&amp;T (biological), and recreational areas (people/use) were highest ranked by partners. The decision-support tool provided an objective method of ranking parcels of land for public outreach efforts by the partners to protect and restore wetland wildlife habitat.

Adaptive Feedback Convolutional‐Neural‐Network‐Based High‐Resolution Reflection‐Waveform Inversion

AbstractFull‐waveform inversion (FWI) applies non‐linear optimization to estimate the velocity model by fitting the observed seismic data. With a smooth starting velocity model, FWI mainly inverts for the shallower background velocity model by fitting the observed direct, diving and refracted data, and updates the interfaces by fitting the observed reflected data. As the deeper parts of background velocity model cannot be effectively updated by fitting the reflected data in FWI, the deeper interfaces are less accurate than the shallower interfaces. To update the deeper background velocity model, many reflection‐waveform inversion (RWI) algorithms were proposed to separate the tomographic and migration components from the reflection‐related gradient. We propose a convolutional‐neural‐network‐based reflection‐waveform inversion (CNN‐RWI) to repeatedly apply the iteratively updated convolutional neural network (CNN) to predict the true velocity model from the smooth starting velocity model (the tomographic components), and the high‐resolution migration image (the migration components). The CNN is iteratively updated by the more representative training data set, which is obtained from the latest CNN‐predicted velocity model by the proposed spatially constrained divisive hierarchical k‐means parcellation method. The more representative the training velocity models are, the more accurate the CNN‐predicted velocity model becomes. Synthetic examples using different portions of the Marmousi2 P‐wave velocity model show that CNN‐RWI inverts for both the shallower and deeper parts of velocity models more accurately than the conjugate‐gradient FWI (CG‐FWI). Both the CNN‐RWI and the CG‐FWI are sensitive to the accuracy of the starting velocity model and the complexity of the unknown true velocity model.

Prediction of Chronological Age in Healthy Elderly Subjects with Machine Learning from MRI Brain Segmentation and Cortical Parcellation

Normal aging is associated with changes in volumetric indices of brain atrophy. A quantitative understanding of age-related brain changes can shed light on successful aging. To investigate the effect of age on global and regional brain volumes and cortical thickness, 3514 magnetic resonance imaging scans were analyzed using automated brain segmentation and parcellation methods in elderly healthy individuals (69–88 years of age). The machine learning algorithm extreme gradient boosting (XGBoost) achieved a mean absolute error of 2 years in predicting the age of new subjects. Feature importance analysis showed that the brain-to-intracranial-volume ratio is the most important feature in predicting age, followed by the hippocampi volumes. The cortical thickness in temporal and parietal lobes showed a superior predictive value than frontal and occipital lobes. Insights from this approach that integrate model prediction and interpretation may help to shorten the current explanatory gap between chronological age and biological brain age.

A coarse-grained parcel method for heat and mass transfer simulations of spray coating processes

The Discrete Element Method (DEM) is commonly used for modeling the flow of particulate materials. Unfortunately, such detailed simulations are computationally very demanding, restricting its use for industrially-scaled processes. The number of particles in a simulation can be reduced by introducing parcels (i.e., “coarse graining”), which – in essence – relies on the increase of the particle diameter for interaction calculations. However, sophisticated models are necessary to preserve the original behavior of the material when using such an approach. Our present contribution extends available coarse-graining concepts by introducing models for (i) particle–fluid mass transfer and (ii) the deposition rate of spray droplets on particles. Our mass transfer model is based on an existing model for heat transfer. For the spray deposition model, we introduce an effective particle diameter to compute the correct amount of droplets that impact particles. We show that these models can be used with confidence up to a coarse-graining level of 5, which we demonstrate for a simple-shaped fluidized bed. The models proposed by us are critical for detailed simulations of spray coating processes since they enable precise particle-droplet-air interaction modeling at low computational cost.

Towards automated in vivo parcellation of the human cerebral cortex using supervised classification of magnetic resonance fingerprinting residuals.

Accurate parcellation of the cerebral cortex in an individual is a guide to its underlying organization. The most promising in vivo quantitative magnetic resonance (MR)-based microstructural cortical mapping methods are yet to achieve a level of parcellation accuracy comparable to quantitative histology. We scanned 6 participants using a 3D echo-planar imaging MR fingerprinting (EPI-MRF) sequence on a 7T Siemens scanner. After projecting MRF signals to the individual-specific inflated model of the cortical surface, normalized autocorrelations of MRF residuals of vertices of 8 microstructurally distinct areas (BA1, BA2, BA4a, BA6, BA44, BA45, BA17, and BA18) from 3 cortical regions were used as feature vector inputs into linear support vector machine (SVM), radial basis function SVM (RBF-SVM), random forest, and k-nearest neighbors supervised classification algorithms. The algorithms' prediction performance was compared using: (i) features from each vertex or (ii) features from neighboring vertices. The neighborhood-based RBF-SVM classifier achieved the highest prediction score of 0.85 for classification of MRF residuals in the central region from a held-out participant. We developed an automated method of cortical parcellation using a combination of MR fingerprinting residual analysis and machine learning classification. Our findings provide the basis for employing unsupervised learning algorithms for whole-cortex structural parcellation in individuals.