Search

How can we predict the occurrence of acute kidney injury (AKI) in cancer patients based on machine learning with serum creatinine data? Given irregular and heterogeneous clinical data, how can we make the most of it for accurate AKI prediction? AKI is a common and significant complication in cancer patients, and correlates with substantial morbidity and mortality. Since no effective treatment for AKI still exists, it is important to take timely preventive measures. While several approaches have been proposed for predicting AKI, their scope and applicability are limited as they either assume regular data measured over a short hospital stay, or do not fully utilize heterogeneous data. In this paper, we provide an AKI prediction model with a greater applicability, which relaxes the constraints of existing approaches, and fully utilizes irregular and heterogeneous data for learning the model. In a cohort of 21,022 cancer patients who were registered into Korea Central Cancer Registry (KCCR) in Seoul National University Hospital between January 1, 2004 and December 31, 2013, our method achieves 0.7892 precision, 0.7506 recall, and 0.7576 F-measure in predicting whether a patient will develop AKI during the next 14 days. [ABSTRACT FROM AUTHOR]

There is a current interest in quantifying time-varying connectivity (TVC) based on neuroimaging data such as fMRI. Many methods have been proposed, and are being applied, revealing new insight into the brain’s dynamics. However, given that the ground truth for TVC in the brain is unknown, many concerns remain regarding the accuracy of proposed estimates. Since there exist many TVC methods it is difficult to assess differences in time-varying connectivity between studies. In this paper, we present tvc_benchmarker, which is a Python package containing four simulations to test TVC methods. Here, we evaluate five different methods that together represent a wide spectrum of current approaches to estimating TVC (sliding window, tapered sliding window, multiplication of temporal derivatives, spatial distance and jackknife correlation). These simulations were designed to test each method’s ability to track changes in covariance over time, which is a key property in TVC analysis. We found that all tested methods correlated positively with each other, but there were large differences in the strength of the correlations between methods. To facilitate comparisons with future TVC methods, we propose that the described simulations can act as benchmark tests for evaluation of methods. Using tvc_benchmarker researchers can easily add, compare and submit their own TVC methods to evaluate its performance. [ABSTRACT FROM AUTHOR]

In this paper we evaluate the three main methods for correcting the susceptibility-induced artefact in diffusion-weighted magnetic-resonance (DW-MR) data, and assess how correction is affected by the susceptibility field’s interaction with motion. The susceptibility artefact adversely impacts analysis performed on the data and is typically corrected in post-processing. Correction strategies involve either registration to a structural image, the application of an acquired field-map or the use of additional images acquired with different phase-encoding. Unfortunately, the choice of which method to use is made difficult by the absence of any systematic comparisons of them. In this work we quantitatively evaluate these methods, by extending and employing a recently proposed framework that allows for the simulation of realistic DW-MR datasets with artefacts. Our analysis separately evaluates the ability for methods to correct for geometric distortions and to recover lost information in regions of signal compression. In terms of geometric distortions, we find that registration-based methods offer the poorest correction. Field-mapping techniques are better, but are influenced by noise and partial volume effects, whilst multiple phase-encode methods performed best. We use our simulations to validate a popular surrogate metric of correction quality, the comparison of corrected data acquired with AP and LR phase-encoding, and apply this surrogate to real datasets. Furthermore, we demonstrate that failing to account for the interaction of the susceptibility field with head movement leads to increased errors when analysing DW-MR data. None of the commonly used post-processing methods account for this interaction, and we suggest this may be a valuable area for future methods development. [ABSTRACT FROM AUTHOR]

Kalman filtering methods have long been regarded as efficient adaptive Bayesian techniques for estimating hidden states in models of linear dynamical systems under Gaussian uncertainty. Recent advents of the Cubature Kalman filter (CKF) have extended this efficient estimation property to nonlinear systems, and also to hybrid nonlinear problems where by the processes are continuous and the observations are discrete (continuous-discrete CD-CKF). Employing CKF techniques, therefore, carries high promise for modeling many biological phenomena where the underlying processes exhibit inherently nonlinear, continuous, and noisy dynamics and the associated measurements are uncertain and time-sampled. This paper investigates the performance of cubature filtering (CKF and CD-CKF) in two flagship problems arising in the field of neuroscience upon relating brain functionality to aggregate neurophysiological recordings: (i) estimation of the firing dynamics and the neural circuit model parameters from electric potentials (EP) observations, and (ii) estimation of the hemodynamic model parameters and the underlying neural drive from BOLD (fMRI) signals. First, in simulated neural circuit models, estimation accuracy was investigated under varying levels of observation noise (SNR), process noise structures, and observation sampling intervals (dt). When compared to the CKF, the CD-CKF consistently exhibited better accuracy for a given SNR, sharp accuracy increase with higher SNR, and persistent error reduction with smaller dt. Remarkably, CD-CKF accuracy shows only a mild deterioration for non-Gaussian process noise, specifically with Poisson noise, a commonly assumed form of background fluctuations in neuronal systems. Second, in simulated hemodynamic models, parametric estimates were consistently improved under CD-CKF. Critically, time-localization of the underlying neural drive, a determinant factor in fMRI-based functional connectivity studies, was significantly more accurate under CD-CKF. In conclusion, and with the CKF recently benchmarked against other advanced Bayesian techniques, the CD-CKF framework could provide significant gains in robustness and accuracy when estimating a variety of biological phenomena models where the underlying process dynamics unfold at time scales faster than those seen in collected measurements. [ABSTRACT FROM AUTHOR]

Detecting changes of spatially high-resolution functional connectivity patterns in the brain is crucial for improving the fundamental understanding of brain function in both health and disease, yet still poses one of the biggest challenges in computational neuroscience. Currently, classical multivariate Granger Causality analyses of directed interactions between single process components in coupled systems are commonly restricted to spatially low- dimensional data, which requires a pre-selection or aggregation of time series as a preprocessing step. In this paper we propose a new fully multivariate Granger Causality approach with embedded dimension reduction that makes it possible to obtain a representation of functional connectivity for spatially high-dimensional data. The resulting functional connectivity networks may consist of several thousand vertices and thus contain more detailed information compared to connectivity networks obtained from approaches based on particular regions of interest. Our large scale Granger Causality approach is applied to synthetic and resting state fMRI data with a focus on how well network community structure, which represents a functional segmentation of the network, is preserved. It is demonstrated that a number of different community detection algorithms, which utilize a variety of algorithmic strategies and exploit topological features differently, reveal meaningful information on the underlying network module structure. [ABSTRACT FROM AUTHOR]

Resting-state functional brain imaging studies of network connectivity have long assumed that functional connections are stationary on the timescale of a typical scan. Interest in moving beyond this simplifying assumption has emerged only recently. The great hope is that training the right lens on time-varying properties of whole-brain network connectivity will shed additional light on previously concealed brain activation patterns characteristic of serious neurological or psychiatric disorders. We present evidence that multiple explicitly dynamical properties of time-varying whole-brain network connectivity are strongly associated with schizophrenia, a complex mental illness whose symptomatic presentation can vary enormously across subjects. As with so much brain-imaging research, a central challenge for dynamic network connectivity lies in determining transformations of the data that both reduce its dimensionality and expose features that are strongly predictive of important population characteristics. Our paper introduces an elegant, simple method of reducing and organizing data around which a large constellation of mutually informative and intuitive dynamical analyses can be performed. This framework combines a discrete multidimensional data-driven representation of connectivity space with four core dynamism measures computed from large-scale properties of each subject’s trajectory, ie., properties not identifiable with any specific moment in time and therefore reasonable to employ in settings lacking inter-subject time-alignment, such as resting-state functional imaging studies. Our analysis exposes pronounced differences between schizophrenia patients (Nsz = 151) and healthy controls (Nhc = 163). Time-varying whole-brain network connectivity patterns are found to be markedly less dynamically active in schizophrenia patients, an effect that is even more pronounced in patients with high levels of hallucinatory behavior. To the best of our knowledge this is the first demonstration that high-level dynamic properties of whole-brain connectivity, generic enough to be commensurable under many decompositions of time-varying connectivity data, exhibit robust and systematic differences between schizophrenia patients and healthy controls. [ABSTRACT FROM AUTHOR]

Over the past several years, significant efforts have been made to improve the spatial resolution of diffusion-weighted imaging (DWI), aiming at better detecting subtle lesions and more reliably resolving white-matter fiber tracts. A major concern with high-resolution DWI is the limited signal-to-noise ratio (SNR), which may significantly offset the advantages of high spatial resolution. Although the SNR of DWI data can be improved by denoising in post-processing, existing denoising procedures may potentially reduce the anatomic resolvability of high-resolution imaging data. Additionally, non-Gaussian noise induced signal bias in low-SNR DWI data may not always be corrected with existing denoising approaches. Here we report an improved denoising procedure, termed diffusion-matched principal component analysis (DM-PCA), which comprises 1) identifying a group of (not necessarily neighboring) voxels that demonstrate very similar magnitude signal variation patterns along the diffusion dimension, 2) correcting low-frequency phase variations in complex-valued DWI data, 3) performing PCA along the diffusion dimension for real- and imaginary-components (in two separate channels) of phase-corrected DWI voxels with matched diffusion properties, 4) suppressing the noisy PCA components in real- and imaginary-components, separately, of phase-corrected DWI data, and 5) combining real- and imaginary-components of denoised DWI data. Our data show that the new two-channel (i.e., for real- and imaginary-components) DM-PCA denoising procedure performs reliably without noticeably compromising anatomic resolvability. Non-Gaussian noise induced signal bias could also be reduced with the new denoising method. The DM-PCA based denoising procedure should prove highly valuable for high-resolution DWI studies in research and clinical uses. [ABSTRACT FROM AUTHOR]

Decision-making is usually accompanied by metacognition, through which a decision maker monitors uncertainty regarding a decision and may then consequently revise the decision. These decisional metacognitive processes can occur prior to or in the absence of feedback. However, the neural mechanisms of metacognition remain controversial. One theory proposes an independent neural system for metacognition in the prefrontal cortex (PFC); the other, that metacognitive processes coincide and overlap with the systems used for the decision-making process per se. In this study, we devised a novel “decision–redecision” paradigm to investigate the neural metacognitive processes involved in redecision as compared to the initial decision-making process. The participants underwent a perceptual decision-making task and a rule-based decision-making task during functional magnetic resonance imaging (fMRI). We found that the anterior PFC, including the dorsal anterior cingulate cortex (dACC) and lateral frontopolar cortex (lFPC), were more extensively activated after the initial decision. The dACC activity in redecision positively scaled with decisional uncertainty and correlated with individual metacognitive uncertainty monitoring abilities—commonly occurring in both tasks—indicating that the dACC was specifically involved in decision uncertainty monitoring. In contrast, the lFPC activity seen in redecision processing was scaled with decision uncertainty reduction and correlated with individual accuracy changes—positively in the rule-based decision-making task and negatively in the perceptual decision-making task. Our results show that the lFPC was specifically involved in metacognitive control of decision adjustment and was subject to different control demands of the tasks. Therefore, our findings support that a separate neural system in the PFC is essentially involved in metacognition and further, that functions of the PFC in metacognition are dissociable. [ABSTRACT FROM AUTHOR]

We propose a novel method to test the existence of community structure in undirected, real-valued, edge-weighted graphs. The method is based on the asymptotic behavior of extreme eigenvalues of a real symmetric edge-weight matrix. We provide a theoretical foundation for this method and report on its performance using synthetic and real data, suggesting that this new method outperforms other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Introduction: The paced serial addition test (PSAT) is regularly used to assess cognitive deficits in various neuropsychiatric conditions. Being a complex test, it reflects the status of multiple cognitive domains such as working memory, information processing speed and executive functioning. Two versions of the PSAT exist. One uses auditory stimuli through spoken numbers and is known as the PASAT, while the other one presents patients with visual stimuli and is called PVSAT. The PASAT is considered more frustrating by patients, and hence the visual version is usually preferred. Research has suggested that an interference might exist between patients’ verbal answers and the auditory presentation of stimuli. We therefore removed the verbal response in this study, and aimed to investigate differences in functional brain activity through functional magnetic resonance imaging. Methods: Fifteen healthy controls performed the two test versions inside an MRI scanner—switching between stimulus modality (auditory vs. visual) as well as inter-stimulus frequency (3s vs. 2s). We extracted 11 independent components from the data: attentional, visual, auditory, sensorimotor and default mode networks. We then performed statistical analyses of mean network activity within each component, as well as inter-network connectivity of each component pair during the different task types. Results: Unsurprisingly, we noted an effect of modality on activity in the visual and auditory components. However, we also describe bilateral frontoparietal, anterior cingulate and insular attentional network activity. An effect of frequency was noted only in the sensorimotor network. Effects were found on edges linking visual and auditory regions. Task modality influenced an attentional-sensorimotor connection, while stimulus frequency had an influence on sensorimotor-default mode connections. Conclusions: Scanner noise during functional MRI may interfere with brain activation—especially during tasks involving auditory pathways. The question whether to use PVSAT or PASAT for an fMRI study is, therefore, an important one. Specific effects of both modalities should be known to study designers. We conclude that both tests should not be considered interchangeable, as significant changes were brought to light during test performance in different modalities. [ABSTRACT FROM AUTHOR]

Resting state functional magnetic resonance imaging (rs-fMRI) in infants enables important studies of functional brain organization early in human development. However, rs-fMRI in infants has universally been obtained during sleep to reduce participant motion artifact, raising the question of whether differences in functional organization between awake adults and sleeping infants that are commonly attributed to development may instead derive, at least in part, from sleep. This question is especially important as rs-fMRI differences in adult wake vs. sleep are well documented. To investigate this question, we compared functional connectivity and BOLD signal propagation patterns in 6, 12, and 24 month old sleeping infants with patterns in adult wakefulness and non-REM sleep. We find that important functional connectivity features seen during infant sleep closely resemble those seen during adult sleep, including reduced default mode network functional connectivity. However, we also find differences between infant and adult sleep, especially in thalamic BOLD signal propagation patterns. These findings highlight the importance of considering sleep state when drawing developmental inferences in infant rs-fMRI. [ABSTRACT FROM AUTHOR]

In diagnostic applications of statistical machine learning methods to brain imaging data, common problems include data high-dimensionality and co-linearity, which often cause over-fitting and instability. To overcome these problems, we applied partial least squares (PLS) regression to resting-state functional magnetic resonance imaging (rs-fMRI) data, creating a low-dimensional representation that relates symptoms to brain activity and that predicts clinical measures. Our experimental results, based upon data from clinically depressed patients and healthy controls, demonstrated that PLS and its kernel variants provided significantly better prediction of clinical measures than ordinary linear regression. Subsequent classification using predicted clinical scores distinguished depressed patients from healthy controls with 80% accuracy. Moreover, loading vectors for latent variables enabled us to identify brain regions relevant to depression, including the default mode network, the right superior frontal gyrus, and the superior motor area. [ABSTRACT FROM AUTHOR]

Objective: The purpose of this study was to investigate the relationship between visual score of emphysema and homology-based emphysema quantification (HEQ) and evaluate whether visual score was accurately predicted by machine learning and HEQ. Materials and methods: A total of 115 anonymized computed tomography images from 39 patients were obtained from a public database. Emphysema quantification of these images was performed by measuring the percentage of low-attenuation lung area (LAA%). The following values related to HEQ were obtained: nb0 and nb1. LAA% and HEQ were calculated at various threshold levels ranging from −1000 HU to −700 HU. Spearman’s correlation coefficients between emphysema quantification and visual score were calculated at the various threshold levels. Visual score was predicted by machine learning and emphysema quantification (LAA% or HEQ). Random Forest was used as a machine learning algorithm, and accuracy of prediction was evaluated by leave-one-patient-out cross validation. The difference in the accuracy was assessed using McNemar’s test. Results: The correlation coefficients between emphysema quantification and visual score were as follows: LAA% (−950 HU), 0.567; LAA% (−910 HU), 0.654; LAA% (−875 HU), 0.704; nb0 (−950 HU), 0.552; nb0 (−910 HU), 0.629; nb0 (−875 HU), 0.473; nb1 (−950 HU), 0.149; nb1 (−910 HU), 0.519; and nb1 (−875 HU), 0.716. The accuracy of prediction was as follows: LAA%, 55.7% and HEQ, 66.1%. The difference in accuracy was statistically significant (p = 0.0290). Conclusion: LAA% and HEQ at −875 HU showed a stronger correlation with visual score than those at −910 or −950 HU. HEQ was more useful than LAA% for predicting visual score. [ABSTRACT FROM AUTHOR]

Objective: To explore the value of machine learning methods for predicting multiple sclerosis disease course. Methods: 1693 CLIMB study patients were classified as increased EDSS≥1.5 (worsening) or not (non-worsening) at up to five years after baseline visit. Support vector machines (SVM) were used to build the classifier, and compared to logistic regression (LR) using demographic, clinical and MRI data obtained at years one and two to predict EDSS at five years follow-up. Results: Baseline data alone provided little predictive value. Clinical observation for one year improved overall SVM sensitivity to 62% and specificity to 65% in predicting worsening cases. The addition of one year MRI data improved sensitivity to 71% and specificity to 68%. Use of non-uniform misclassification costs in the SVM model, weighting towards increased sensitivity, improved predictions (up to 86%). Sensitivity, specificity, and overall accuracy improved minimally with additional follow-up data. Predictions improved within specific groups defined by baseline EDSS. LR performed more poorly than SVM in most cases. Race, family history of MS, and brain parenchymal fraction, ranked highly as predictors of the non-worsening group. Brain T2 lesion volume ranked highly as predictive of the worsening group. Interpretation: SVM incorporating short-term clinical and brain MRI data, class imbalance corrective measures, and classification costs may be a promising means to predict MS disease course, and for selection of patients suitable for more aggressive treatment regimens. [ABSTRACT FROM AUTHOR]

Artifact removal in resting state fMRI (rfMRI) data remains a serious challenge, with even subtle head motion undermining reliability and reproducibility. Here we compared some of the most popular single-echo de-noising methods—regression of Motion parameters, White matter and Cerebrospinal fluid signals (MWC method), FMRIB’s ICA-based X-noiseifier (FIX) and ICA-based Automatic Removal Of Motion Artifacts (ICA-AROMA)—with a multi-echo approach (ME-ICA) that exploits the linear dependency of BOLD on the echo time. Data were acquired using a clinical scanner and included 30 young, healthy participants (minimal head motion) and 30 Attention Deficit Hyperactivity Disorder patients (greater head motion). De-noising effectiveness was assessed in terms of data quality after each cleanup procedure, ability to uncouple BOLD signal and motion and preservation of default mode network (DMN) functional connectivity. Most cleaning methods showed a positive impact on data quality. However, based on the investigated metrics, ME-ICA was the most robust. It minimized the impact of motion on FC even for high motion participants and preserved DMN functional connectivity structure. The high-quality results obtained using ME-ICA suggest that using a multi-echo EPI sequence, reliable rfMRI data can be obtained in a clinical setting. [ABSTRACT FROM AUTHOR]

Independent component analysis (ICA) is widely used in the field of functional neuroimaging to decompose data into spatio-temporal patterns of co-activation. In particular, ICA has found wide usage in the analysis of resting state fMRI (rs-fMRI) data. Recently, a number of large-scale data sets have become publicly available that consist of rs-fMRI scans from thousands of subjects. As a result, efficient ICA algorithms that scale well to the increased number of subjects are required. To address this problem, we propose a two-stage likelihood-based algorithm for performing group ICA, which we denote Parallel Group Independent Component Analysis (PGICA). By utilizing the sequential nature of the algorithm and parallel computing techniques, we are able to efficiently analyze data sets from large numbers of subjects. We illustrate the efficacy of PGICA, which has been implemented in R and is freely available through the Comprehensive R Archive Network, through simulation studies and application to rs-fMRI data from two large multi-subject data sets, consisting of 301 and 779 subjects respectively. [ABSTRACT FROM AUTHOR]

Brain oscillations and synchronicity among brain regions (brain connectivity) have been studied in resting-state (RS) and task-induced settings. RS-connectivity which captures brain functional integration during an unconstrained state is shown to vary with the frequency of oscillations. Indeed, high temporal resolution modalities have demonstrated both between and cross-frequency connectivity spanning across frequency bands such as theta and gamma. Despite high spatial resolution, functional magnetic resonance imaging (fMRI) suffers from low temporal resolution due to modulation with slow-varying hemodynamic response function (HRF) and also relatively low sampling rate. This limits the range of detectable frequency bands in fMRI and consequently there has been no evidence of cross-frequency dependence in fMRI data. In the present work we uncover recurring patterns of spectral power in network timecourses which provides new insight on the actual nature of frequency variation in fMRI network activations. Moreover, we introduce a new measure of dependence between pairs of rs-fMRI networks which reveals significant cross-frequency dependence between functional brain networks specifically default-mode, cerebellar and visual networks. This is the first strong evidence of cross-frequency dependence between functional networks in fMRI and our subject group analysis based on age and gender supports usefulness of this observation for future clinical applications. [ABSTRACT FROM AUTHOR]

Objective: We aimed to identify the consistent regions of gray matter volume (GMV) abnormalities in idiopathic generalized epilepsy (IGE), and to study the difference of GMV abnormalities among IGE subsyndromes by applying activation likelihood estimation (ALE) meta-analysis. Methods: A systematic review of VBM studies on GMV of patients with absence epilepsy (AE), juvenile myoclonic epilepsy (JME), IGE and controls indexed in PubMed and ScienceDirect from January 1999 to June 2016 was conducted. A total of 12 IGE studies, including 7 JME and 3 AE studies, were selected. Meta-analysis was performed on these studies by using the pooled and within-subtypes analysis (). Based on the above results, between-subtypes contrast analysis was carried out to detect the abnormal GMV regions common in and unique to each subtype as well. Results: IGE demonstrated significant GMV increase in right ventral lateral nucleus (VL) and right medial frontal gyrus, and significant GMV decrease in bilateral pulvinar. For JME, significant GMV increase was seen in right medial frontal gyrus, right anterior cingulate cortex (ACC), while significant GMV decrease was found in right pulvinar. In AE, the most significant GMV increase was found in right VL, and slight GMV reduction was seen in right medial dorsal nucleus, right subcallosal gyrus, left caudate and left precuneus. No overlapped and unique regions with significant GMV abnormalities were found between JME and AE. Significance: This meta-analysis demonstrated that thalamo-frontal network was a structure with significant GMV abnormality in IGE, and the IGE subsyndromes showed different GMV abnormal regions. These observations may provide instructions on the clinical diagnosis of IGE. [ABSTRACT FROM AUTHOR]

Individual differences in brain functional networks may be related to complex personal identifiers, including health, age, and ability. Dynamic network theory has been used to identify properties of dynamic brain function from fMRI data, but the majority of analyses and findings remain at the level of the group. Here, we apply hypergraph analysis, a method from dynamic network theory, to quantify individual differences in brain functional dynamics. Using a summary metric derived from the hypergraph formalism—hypergraph cardinality—we investigate individual variations in two separate, complementary data sets. The first data set (“multi-task”) consists of 77 individuals engaging in four consecutive cognitive tasks. We observe that hypergraph cardinality exhibits variation across individuals while remaining consistent within individuals between tasks; moreover, the analysis of one of the memory tasks revealed a marginally significant correspondence between hypergraph cardinality and age. This finding motivated a similar analysis of the second data set (“age-memory”), in which 95 individuals, aged 18–75, performed a memory task with a similar structure to the multi-task memory task. With the increased age range in the age-memory data set, the correlation between hypergraph cardinality and age correspondence becomes significant. We discuss these results in the context of the well-known finding linking age with network structure, and suggest that hypergraph analysis should serve as a useful tool in furthering our understanding of the dynamic network structure of the brain. [ABSTRACT FROM AUTHOR]

Breast screening is an important tool for the early detection of breast cancers. However, tumours are typically present in less than 1% of mammograms. This low prevalence could cause radiologists to detect fewer tumours than they otherwise would, an issue known as the prevalence effect. The aim of our study was to investigate a novel breast screening protocol, designed to decrease the number of tumours missed by radiologists, without increasing their workload. We ran two laboratory-based experiments to assess the degree to which the novel protocol, called the catch trial (CT) protocol, resulted in greater sensitivity (d’) than the double screener protocol (DS), currently utilised in Australia. In our first experiment we found evidence that the CT protocol resulted in a criterion shift relative to the DS protocol but the evidence that sensitivity was greater in the CT protocol relative to the DS protocol was less clear. A second experiment, using more realistic stimuli that were more representative of actual tumours, also failed to find convincing evidence that sensitivity was greater in the CT protocol than in the DS protocol. This experiment instead found that both the hit rate and the false alarm rate increased in the CT protocol relative to the DS protocol. So while there was again evidence that the CT protocol induced a criterion shift, the sensitivity appeared to be approximately the same in both protocols. Our results suggest the CT protocol is unlikely to result in an improvement in sensitivity over the DS protocol, so we cannot recommend that it be trialled in a clinical setting. [ABSTRACT FROM AUTHOR]

In this research we have analyzed functional magnetic resonance imaging (fMRI) signals of different networks in the brain under resting state condition. To such end, the dynamics of signal variation, have been conceived as a stochastic motion, namely it has been modelled through a generalized Langevin stochastic differential equation, which combines a deterministic drift component with a stochastic component where the Gaussian noise source has been replaced with α-stable noise. The parameters of the deterministic and stochastic parts of the model have been fitted from fluctuating data. Results show that the deterministic part is characterized by a simple, linear decreasing trend, and, most important, the α-stable noise, at varying characteristic index α, is the source of a spectrum of activity modes across the networks, from those originated by classic Gaussian noise (α = 2), to longer tailed behaviors generated by the more general Lévy noise (1 ≤ α < 2). Lévy motion is a specific instance of scale-free behavior, it is a source of anomalous diffusion and it has been related to many aspects of human cognition, such as information foraging through memory retrieval or visual exploration. Finally, some conclusions have been drawn on the functional significance of the dynamics corresponding to different α values. [ABSTRACT FROM AUTHOR]

The classification of neuroimaging data for the diagnosis of certain brain diseases is one of the main research goals of the neuroscience and clinical communities. In this study, we performed multiclass classification using a hierarchical extreme learning machine (H-ELM) classifier. We compared the performance of this classifier with that of a support vector machine (SVM) and basic extreme learning machine (ELM) for cortical MRI data from attention deficit/hyperactivity disorder (ADHD) patients. We used 159 structural MRI images of children from the publicly available ADHD-200 MRI dataset. The data consisted of three types, namely, typically developing (TDC), ADHD-inattentive (ADHD-I), and ADHD-combined (ADHD-C). We carried out feature selection by using standard SVM-based recursive feature elimination (RFE-SVM) that enabled us to achieve good classification accuracy (60.78%). In this study, we found the RFE-SVM feature selection approach in combination with H-ELM to effectively enable the acquisition of high multiclass classification accuracy rates for structural neuroimaging data. In addition, we found that the most important features for classification were the surface area of the superior frontal lobe, and the cortical thickness, volume, and mean surface area of the whole cortex. [ABSTRACT FROM AUTHOR]

Background: Traumatic brain injury remains a global health problem. Understanding the relative importance of outcome predictors helps optimize our treatment strategies by informing assessment protocols, clinical decisions and trial designs. In this study we establish importance ranking for outcome predictors based on receiver operating indices to identify key predictors of outcome and create simple predictive models. We then explore the associations between key outcome predictors using Bayesian networks to gain further insight into predictor importance. Methods: We analyzed the corticosteroid randomization after significant head injury (CRASH) trial database of 10008 patients and included patients for whom demographics, injury characteristics, computer tomography (CT) findings and Glasgow Outcome Scale (GCS) were recorded (total of 13 predictors, which would be available to clinicians within a few hours following the injury in 6945 patients). Predictions of clinical outcome (death or severe disability at 6 months) were performed using logistic regression models with 5-fold cross validation. Predictive performance was measured using standardized partial area (pAUC) under the receiver operating curve (ROC) and we used Delong test for comparisons. Variable importance ranking was based on pAUC targeted at specificity (pAUCSP) and sensitivity (pAUCSE) intervals of 90–100%. Probabilistic associations were depicted using Bayesian networks. Results: Complete AUC analysis showed very good predictive power (AUC = 0.8237, 95% CI: 0.8138–0.8336) for the complete model. Specificity focused importance ranking highlighted age, pupillary, motor responses, obliteration of basal cisterns/3rd ventricle and midline shift. Interestingly when targeting model sensitivity, the highest-ranking variables were age, severe extracranial injury, verbal response, hematoma on CT and motor response. Simplified models, which included only these key predictors, had similar performance (pAUCSP = 0.6523, 95% CI: 0.6402–0.6641 and pAUCSE = 0.6332, 95% CI: 0.62–0.6477) compared to the complete models (pAUCSP = 0.6664, 95% CI: 0.6543–0.679, pAUCSE = 0.6436, 95% CI: 0.6289–0.6585, de Long p value 0.1165 and 0.3448 respectively). Bayesian networks showed the predictors that did not feature in the simplified models were associated with those that did. Conclusion: We demonstrate that importance based variable selection allows simplified predictive models to be created while maintaining prediction accuracy. Variable selection targeting specificity confirmed key components of clinical assessment in TBI whereas sensitivity based ranking suggested extracranial injury as one of the important predictors. These results help refine our approach to head injury assessment, decision-making and outcome prediction targeted at model sensitivity and specificity. Bayesian networks proved to be a comprehensive tool for depicting probabilistic associations for key predictors giving insight into why the simplified model has maintained accuracy. [ABSTRACT FROM AUTHOR]

The human brain can be parcellated into diverse anatomical areas. We investigated whether rhythmic brain activity in these areas is characteristic and can be used for automatic classification. To this end, resting-state MEG data of 22 healthy adults was analysed. Power spectra of 1-s long data segments for atlas-defined brain areas were clustered into spectral profiles (“fingerprints”), using k-means and Gaussian mixture (GM) modelling. We demonstrate that individual areas can be identified from these spectral profiles with high accuracy. Our results suggest that each brain area engages in different spectral modes that are characteristic for individual areas. Clustering of brain areas according to similarity of spectral profiles reveals well-known brain networks. Furthermore, we demonstrate task-specific modulations of auditory spectral profiles during auditory processing. These findings have important implications for the classification of regional spectral activity and allow for novel approaches in neuroimaging and neurostimulation in health and disease. [ABSTRACT FROM AUTHOR]

The formation of transient networks in response to external stimuli or as a reflection of internal cognitive processes is a hallmark of human brain function. However, its identification in fMRI data of the human brain is notoriously difficult. Here we propose a new method of fMRI data analysis that tackles this problem by considering large-scale, task-related synchronisation networks. Networks consist of nodes and edges connecting them, where nodes correspond to voxels in fMRI data, and the weight of an edge is determined via task-related changes in dynamic synchronisation between their respective times series. Based on these definitions, we developed a new data analysis algorithm that identifies edges that show differing levels of synchrony between two distinct task conditions and that occur in dense packs with similar characteristics. Hence, we call this approach “Task-related Edge Density” (TED). TED proved to be a very strong marker for dynamic network formation that easily lends itself to statistical analysis using large scale statistical inference. A major advantage of TED compared to other methods is that it does not depend on any specific hemodynamic response model, and it also does not require a presegmentation of the data for dimensionality reduction as it can handle large networks consisting of tens of thousands of voxels. We applied TED to fMRI data of a fingertapping and an emotion processing task provided by the Human Connectome Project. TED revealed network-based involvement of a large number of brain areas that evaded detection using traditional GLM-based analysis. We show that our proposed method provides an entirely new window into the immense complexity of human brain function. [ABSTRACT FROM AUTHOR]

Systems neuroscience has identified a set of canonical large-scale networks in humans. These have predominantly been characterized by resting-state analyses of the task-unconstrained, mind-wandering brain. Their explicit relationship to defined task performance is largely unknown and remains challenging. The present work contributes a multivariate statistical learning approach that can extract the major brain networks and quantify their configuration during various psychological tasks. The method is validated in two extensive datasets (n = 500 and n = 81) by model-based generation of synthetic activity maps from recombination of shared network topographies. To study a use case, we formally revisited the poorly understood difference between neural activity underlying idling versus goal-directed behavior. We demonstrate that task-specific neural activity patterns can be explained by plausible combinations of resting-state networks. The possibility of decomposing a mental task into the relative contributions of major brain networks, the "network co-occurrence architecture" of a given task, opens an alternative access to the neural substrates of human cognition. [ABSTRACT FROM AUTHOR]

Echo planar imaging (EPI) suffers from geometric distortions caused by magnetic field inhomogeneities, which can be time-varying as a result of small amounts of head motion that occur over seconds and minutes during fMRI experiments, also known as “dynamic geometric distortion”. Phase Labeling for Additional Coordinate Encoding (PLACE) is a promising technique for geometric distortion correction without reduced temporal resolution and in principle can be used to correct for motion-induced dynamic geometric distortion. PLACE requires at least two EPI images of the same anatomy that are ideally acquired with no variation in the magnetic field inhomogeneities. However, head motion and lung ventilation during the respiratory cycle can cause changes in magnetic field inhomogeneities within the EPI pair used for PLACE. In this work, we exploited dynamic off-resonance in k-space (DORK) and averaging to correct the within EPI pair magnetic field inhomogeneities; and hence proposed a combined technique (DORK+PLACE+averaging) to mitigate dynamic geometric distortion in EPI-based fMRI while preserving the temporal resolution. The performance of the combined DORK, PLACE and averaging technique was characterized through several imaging experiments involving test phantoms and six healthy adult volunteers. Phantom data illustrate reduced temporal standard deviation of fMRI signal intensities after use of combined dynamic PLACE, DORK and averaging compared to the standard processing and static geometric distortion correction. The combined technique also substantially improved the temporal standard deviation and activation maps obtained from human fMRI data in comparison to the results obtained by standard processing and static geometric distortion correction, highlighting the utility of the approach. [ABSTRACT FROM AUTHOR]

We compared the efficacy of contrast-enhanced ultrasound (CEUS) and contrast-enhanced computed tomography (CECT) for the diagnosis of renal cystic lesions via a meta-analysis to determine the value of CEUS in the prediction of the malignant potential of complex renal cysts. Eleven studies were evaluated: 4 control studies related to CEUS and CECT, 3 studies related to CEUS and 4 studies related to CECT. According to the random effects model, the pooled sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio for CEUS/CECT were 0.95/0.90, 0.79/0.85, 4.39/5.00, and 0.10/0.15, respectively. The areas under the summary receiver operating characteristic (AUCs-SROC) curves for the two methods were 94.24% and 93.39%, and the estimated Q values were 0.8805 and 0.8698, respectively. Comparing the Q index values of CEUS and CECT revealed no significant difference between the two methods (P>0.05). When compared with conventional CECT, CEUS is also useful for diagnosing renal cystic lesions in the clinic. [ABSTRACT FROM AUTHOR]

Background: Preprocedural manual multi-slice-CT-segmentation tools (MSCT-ST) define the gold standard for planning transcatheter aortic valve replacement (TAVR). They are able to predict the perpendicular line of the aortic annulus (PPL) and to indicate the corresponding C-arm angulation (CAA). Fully automated planning-tools and their clinical relevance have not been systematically evaluated in a real world setting so far. Methods and Results: The study population consists of an all-comers cohort of 160 consecutive TAVR patients with a drop out of 35 patients for technical and anatomical reasons. 125 TAVR patients underwent preprocedural analysis by manual (M-MSCT) and fully automated MSCT-ST (A-MSCT). Method-comparison was performed for 105 patients (Cohort A). In Cohort A, CAA was defined for each patient, and accordance within 10° between M-MSCT and A-MSCT was considered adequate for concept-proof (95% in LAO/RAO; 94% in CRAN/CAUD). Intraprocedural CAA was defined by repetitive angiograms without utilizing the preprocedural measurements. In Cohort B, intraprocedural CAA was established with the use of A-MSCT (20 patients). Using preprocedural A-MSCT to indicate the corresponding CAA, the levels of contrast medium (ml) and radiation exposure (cine runs) were reduced in Cohort B compared to Cohort A significantly (23.3±10.3 vs. 35.3 ±21.1 ml, p = 0.02; 1.6±0.7 vs. 2.4±1.4 cine runs; p = 0.02) and trends towards more safety in valve-positioning could be demonstrated. Conclusions: A-MSCT-analysis provides precise preprocedural information on CAA for optimal visualization of the aortic annulus compared to the M-MSCT gold standard. Intraprocedural application of this information during TAVR significantly reduces the levels of contrast and radiation exposure. Trial Registration: ClinicalTrials.gov [ABSTRACT FROM AUTHOR]

We present a complexity-based approach for the analysis of fMRI time series, in which sample entropy (SampEn) is introduced as a quantification of the voxel complexity. Under this hypothesis the voxel complexity could be modulated in pertinent cognitive tasks, and it changes through experimental paradigms. We calculate the complexity of sequential fMRI data for each voxel in two distinct experimental paradigms and use a nonparametric statistical strategy, the Wilcoxon signed rank test, to evaluate the difference in complexity between them. The results are compared with the well known general linear model based Statistical Parametric Mapping package (SPM12), where a decided difference has been observed. This is because SampEn method detects brain complexity changes in two experiments of different conditions and the data-driven method SampEn evaluates just the complexity of specific sequential fMRI data. Also, the larger and smaller SampEn values correspond to different meanings, and the neutral-blank design produces higher predictability than threat-neutral. Complexity information can be considered as a complementary method to the existing fMRI analysis strategies, and it may help improving the understanding of human brain functions from a different perspective. [ABSTRACT FROM AUTHOR]

Signal detection in functional magnetic resonance imaging (fMRI) inherently involves the problem of testing a large number of hypotheses. A popular strategy to address this multiplicity is the control of the false discovery rate (FDR). In this work we consider the case where prior knowledge is available to partition the set of all hypotheses into disjoint subsets or families, e. g., by a-priori knowledge on the functionality of certain regions of interest. If the proportion of true null hypotheses differs between families, this structural information can be used to increase statistical power. We propose a two-stage multiple test procedure which first excludes those families from the analysis for which there is no strong evidence for containing true alternatives. We show control of the family-wise error rate at this first stage of testing. Then, at the second stage, we proceed to test the hypotheses within each non-excluded family and obtain asymptotic control of the FDR within each family at this second stage. Our main mathematical result is that this two-stage strategy implies asymptotic control of the FDR with respect to all hypotheses. In simulations we demonstrate the increased power of this new procedure in comparison with established procedures in situations with highly unbalanced families. Finally, we apply the proposed method to simulated and to real fMRI data. [ABSTRACT FROM AUTHOR]

Functional Magnetic Resonance Imaging (fMRI) based brain connectivity analysis maps the functional networks of the brain by estimating the degree of synchronous neuronal activity between brain regions. Recent studies have demonstrated that “resting-state” fMRI-based brain connectivity conclusions may be erroneous when motion artifacts have a differential effect on fMRI BOLD signals for between group comparisons. A potential explanation could be that in-scanner displacement, due to rotational components, is not spatially constant in the whole brain. However, this localized nature of motion artifacts is poorly understood and is rarely considered in brain connectivity studies. In this study, we initially demonstrate the local correspondence between head displacement and the changes in the resting-state fMRI BOLD signal. Than, we investigate how connectivity strength is affected by the population-level variation in the spatial pattern of regional displacement. We introduce Regional Displacement Interaction (RDI), a new covariate parameter set for second-level connectivity analysis and demonstrate its effectiveness in reducing motion related confounds in comparisons of groups with different voxel-vise displacement pattern and preprocessed using various nuisance regression methods. The effect of using RDI as second-level covariate is than demonstrated in autism-related group comparisons. The relationship between the proposed method and some of the prevailing subject-level nuisance regression techniques is evaluated. Our results show that, depending on experimental design, treating in-scanner head motion as a global confound may not be appropriate. The degree of displacement is highly variable among various brain regions, both within and between subjects. These regional differences bias correlation-based measures of brain connectivity. The inclusion of the proposed second-level covariate into the analysis successfully reduces artifactual motion-related group differences and preserves real neuronal differences, as demonstrated by the autism-related comparisons. [ABSTRACT FROM AUTHOR]

Phase-Contrast MRI (PC-MRI) is a noninvasive technique to measure blood flow. In particular, global but highly quantitative cerebral blood flow (CBF) measurement using PC-MRI complements several other CBF mapping methods such as arterial spin labeling and dynamic susceptibility contrast MRI by providing a calibration factor. The ability to estimate blood supply in physiological units also lays a foundation for assessment of brain metabolic rate. However, a major obstacle before wider applications of this method is that the slice positioning of the scan, ideally placed perpendicular to the feeding arteries, requires considerable expertise and can present a burden to the operator. In the present work, we proposed that the majority of PC-MRI scans can be positioned using an automatic algorithm, leaving only a small fraction of arteries requiring manual positioning. We implemented and evaluated an algorithm for this purpose based on feature extraction of a survey angiogram, which is of minimal operator dependence. In a comparative test-retest study with 7 subjects, the blood flow measurement using this algorithm showed an inter-session coefficient of variation (CoV) of . The Bland-Altman method showed that the automatic method differs from the manual method by between and , for of the CBF measurements. This is comparable to the variance in CBF measurement using manually-positioned PC MRI alone. In a further application of this algorithm to 157 consecutive subjects from typical clinical cohorts, the algorithm provided successful positioning in 89.7% of the arteries. In 79.6% of the subjects, all four arteries could be planned using the algorithm. Chi-square tests of independence showed that the success rate was not dependent on the age or gender, but the patients showed a trend of lower success rate (p = 0.14) compared to healthy controls. In conclusion, this automatic positioning algorithm could improve the application of PC-MRI in CBF quantification. [ABSTRACT FROM AUTHOR]