Hemodynamic Response Function (HRF)

PyHRF PyHRF ( http://www.pyhrf.org ) is an open source tool imple- mented in Python that allows to jointly detect activation and estimate (JDE) the hemodynamic response function (HRF) [ MIV + 08 ], which gives the temporal changes in the BOLD effect induced by brain activity. This estimation is not easy in a voxel- wise manner [ PJG + 03 ], and a spatial structure was added to JDE [ VRC10 ] in order to get reliable estimates. In this regard, HRF estimation in JDE is parcel-wise and an input parcellation is required. However, the use of the Markov Chain Monte Carlo

Figure 6: statistics on pseudoevents and resting HRF parameters th a contrast following the level of consciousness, and typical shapes within the cluster (bottom right) CONCLUSIONS AND FUTURE WORK We have presented a methodology to retrieve the hemodynamic response function at from resting state functional magnetic resonance imaging data. The results are promising since the shape of the retrieved HRF is consistent with the literature and supports evidences of the vascular flow. Additionally, the functional modifications to the HRF shape are consistent with evidence previously reported using different methodologies. The approach will need further validation using

a and b: comparison of the HRF estimates computed from one vs four sessions, respectively when the low frequency drift included in the data is modeled with &Û/Ý nuisance variables for ea[r]

b INRIA, MISTIS, Grenoble University, LJK, Grenoble, France c CEA/NeuroSpin and INRIA Saclay, Parietal, France Abstract One of the most challenging issues in task-related fMRI data analysis con- sists of deriving a meaningful functional brain parcellation. The joint par- cellation detection estimation (JPDE) model addresses this issue through an automatic inference of the parcels directly from fMRI data. However, for doing so, the number of parcels needs to be fixed a priori and an appropri- ate initialization for the mask parcellation must be provided too. Hence, this difficult task generally depends on the subject. In this paper, an auto- matic model selection approach is proposed to overcome this limitation at the subject-level. Our approach relies on a non-parametric Bayesian approach that estimates the number of parcels online using a Dirichlet process mixture model combined with a hidden Markov random field. The inference is carried out using a variational expectation maximization strategy. As compared to a standard model selection approach in the original JPDE framework, our non-parametric extension appears more efficient in terms of computational time and does not require finely tuned initialization. Our method is first validated on synthetic data to demonstrate its robustness in selecting the right model order and providing accurate estimates for the parcellation, the hemodynamic response function (HRF) shapes and the activation maps. The method is then validated on real fMRI data in two regions of interest (ROIs): right motor and bilateral occipital ROIs. The results show the ability of the proposed method to aggregate parcels with similar behaviour from a hemo- dynamic point of view, while discriminating them from other parcels having

Guillaume Marrelec*, Philippe Ciuciu, Member, IEEE, Mélanie Pélégrini-Issac, and Habib Benali, Senior Member, IEEE Abstract—A convenient way to analyze blood-oxygen-level-de- pendent functional magnetic resonance imaging data consists of modeling the whole brain as a stationary, linear system char- acterized by its transfer function: the hemodynamic response function (HRF). HRF estimation, though of the greatest interest, is still under investigation, for the problem is ill-conditioned. In this paper, we recall the most general Bayesian model for HRF estimation and show how it can beneficially be translated in terms of Bayesian graphical models, leading to 1) a clear and efficient representation of all structural and functional relationships entailed by the model, and 2) a straightforward numerical scheme to approximate the joint posterior distribution, allowing for estimation of the HRF, as well as all other model parameters. We finally apply this novel technique on both simulations and real data.

Hemoglobin concentration changes were filtered with a Gauss- ian kernel (1.5 s FWHM) and high pass filtered by a second-order Butterworth filter with a cutoff frequency of 0.01 Hz. The signif- icance of each effect of interest (abstract, concrete, and pseudo- word) was determined using the theory of Gaussian fields ( 47 ). A GLM was fit using a canonical hemodynamic response function (HRF). Contrasts over sessions (intrasubject) were analyzed using a fixed effects model, while testing for contrasts in the intersubject analysis was done by estimating the ratio of the random effects variance to the fixed effects variance. An expected Euler correction based on Lipschitz–Killing curvatures was applied to the threshold on the HbR/HbO2 t -statistic images to account for the spatial cor- relation. The GLM method was based on the precoloring method of NIRS-SPM toolbox ( 51 ) for noise treatment.

Marseille, France, 3 Centre National de la Recherche Scientifique, Centre IRM Fonctionnelle Cérébrale, Institut de Neurosciences de la Timone UMR 7289, Aix-Marseille Université, Marseille, France Conventional analysis of functional magnetic resonance imaging (fMRI) data using the general linear model (GLM) employs a neural model convolved with a canonical hemodynamic response function (HRF) peaking 5 s after stimulation. Incorporation of a further basis function, namely the canonical HRF temporal derivative, accounts for delays in the hemodynamic response to neural activity. A population that may benefit from this flexible approach is children whose hemodynamic response is not yet mature. Here, we examined the effects of using the set based on the canonical HRF plus its temporal derivative on both first- and second-level GLM analyses, through simulations and using developmental data (an fMRI dataset on proprioceptive mapping in children and adults). Simulations of delayed fMRI first-level data emphasized the benefit of carrying forward to the second-level a derivative boost that combines derivative and nonderivative beta estimates. In the experimental data, second-level analysis using a paired t-test showed increased mean amplitude estimate (i.e., increased group contrast mean) in several brain regions related to proprioceptive processing when using the derivative boost compared to using only the nonderivative term. This was true especially in children. However, carrying forward to the second-level the individual derivative boosts had adverse consequences on random-effects analysis that implemented one-sample t-test, yielding increased between-subject variance, thus affecting group-level statistic. Boosted data also presented a lower level of smoothness that had implication for the detection of group average activation. Imposing soft constraints on the derivative boost by limiting the time-to-peak range of the modeled response within a specified range (i.e., 4–6 s) mitigated these issues. These findings support the notion that there are pros and cons to using the informed basis set with developmental data.

Abstract—Modern cognitive experiments in functional Mag- netic Resonance Imaging (fMRI) often aim at understanding the temporal dynamics of the brain response in regions acti- vated by a given stimulus. The study of the variability of the hemodynamic response function (HRF) and its characteristics can provide some answers. In this context, we aim at improving the accuracy of the HRF estimation. To do so, we relied on a Joint-Detection-Estimation (JDE) framework that enables robust detection of brain activity as well as HRF estimation, in a Bayesian setting [2]. So far, the hemodynamic results provided by the JDE formalism have depended on a prior parcellation of the data performed before JDE inference. In this study, we propose a new approach to relax this prior knowledge: using consensus clustering techniques based on random parcellations of the data, we combine hemodynamics results provided by different parcellations, so as to robustify the HRF estimation. Keywords-fMRI; Consensus Clustering; Random parcella- tion; Hemodynamic estimation; Bayesian inference

extinction curves. The obtained values reflect relative changes of HbO and HbR chromophores in the brain. GLM analyses were performed in the standard manner (Cohen-Adad et al., 2007; Machado et al., 2011; Peng et al., 2013; Pouliot et al., 2012; Ye et al., 2009) by convolving the onset times of epileptic spikes and seizures with a hemodynamic response function (HRF). When spikes arise in a temporally clustered way, nonlinearities could be important (Pouliot et al., 2012)). Thus the leading order nonlinear effect was included as an additional regressor in the GLM, for patients satisfying the rule of thumb that spikes occurred at a rate greater than 100 spikes per 15 minute recording session. A partial component analysis (PCA) filter was performed, consisting in setting to zero the 5 largest eigenvalues in the decomposition over channels. This was observed in (Peng et al., 2013) to yield similar results to setting only the largest eigenvalue to zero. Either of the two PCA filters led to more congruent results than not performing a PCA at all. Temporal filters were also applied: by convolving the data with the HRF for the low pass and applying an order 2 Butterworth at 0.01 Hz for the high pass. Assuming a 95% confidence level, the least stringent of the Euler Characteristic (EC) (Li et al., 2012) and Bonferroni thresholds, as well as a weaker threshold of 3.0 (chosen as a proxy to a peak false discovery rate (FDR) threshold, and corresponding to an uncorrected threshold of 0.00667), were considered as significant thresholds on t- statistics.

1 Introduction Functional magnetic resonance imaging (fMRI) is a powerful non-invasive imag- ining technique to indirectly measure neural activity from the blood-oxygen-level dependent (BOLD) signal [1]. fMRI data analysis relies on two main task; the detection of activation in brain areas after a given stimulus and the estimation of the underlying dynamics of the brain which is also called as the hemodynamic response function (HRF). Many attempts to describe the link between stimuli and the induced BOLD signal have been proposed. The basic model is the so- called general linear model (GLM). In this model, the link between the stimuli and the induced BOLD signal is modelled as a convolution between the HRF and the binary stimulus sequence. However, this model considers a fixed HRF shape [2,3]. Many extensions have been proposed to account for the variability of

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignemen[r]

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignemen[r]

, and Maha Ayyoub 1 . 4 . s Although understanding of T cell exhaustion is widely based on mouse models, its analysis in patients with cancer could provide clues indicating tumor sensitivity to immune checkpoint blockade (ICB). Data suggest a role for costimulatory pathways, particularly CD28, in exhausted T cell responsiveness to PD 1/PD L1 blockade. Here, we used single cell transcriptomic, phenotypic, and function al approaches to dissect the relation between cos + T cell exhaus tion, CD28 costimulation, and tumor specificity in head and neck, cervical, and ovarian cancers. We found that memory tumor specific cos + T cells, but not bystander cells, sequentially express immune checkpoints once they infiltrate tumors, leading, in situ, to a functionally exhausted population. Exhausted T cells were none Introduction

Pretreatment tumor biopsies used for IHC analyses were obtained from a second cohort of patients with head and neck squamous cell carcinoma receiving ICB therapy with PD 1/PD L1 blocking agents (nivolumab, n 21; pembrolizumab, n 1; durvalumab, n 7; and avelumab, n 1) to treat locally advanced or metastatic disease. Biopsies were either obtained up to 2 months prior to the first ICB dosing (5 patients) or retrieved from archival samples (25 patients: <1 year, 8 patients; ≥1 year and <2 years, 9 patients; ≥2 years and <3 years, 4 patients; and ≥3 years and <5 years, 4 patients). Samples were handled by the Biopathological Support Platform for Clinical Studies, IUCT O. Response to therapy was evaluated by iRECIST criteria. Progressive disease (PD) was de fined as the increase of >20% of target lesions or appearance and increase in size of new lesions in at least two CT scan evaluations performed at least 4 weeks apart. Partial response (PR) was de fined as a decrease of >30% in target lesions and complete response (CR) as disappearance of target and non target lesions, both in at least two CT scans performed at least 4 weeks apart. Any response other than PD or PR/CR was con sidered as stable disease (SD).

126 avoiding the middle FF input layer(Felleman and Van Essen 1991) and originate from nearby visual areas (e.g. V2-V4) as well as distant non-visual areas including parrahippocampal gyrus and inferotemporal cortex(Rockland and Van Hoesen n.d.; Kobayashi and Amaral 2003, 2007; Clavagnier et al. 2004; Whittingstall et al. 2014; Mejias et al. 2016). Put otherwise, FB projections to V1 target over twice as many neurons and glia compared to FF (Giannaris and Rosene 2012), indicating that the metabolic burden of integrating and processing even weak FB input is likely to increase hemodynamic measures. Indeed, general anesthesia, which has been shown to selectively disrupt FB while leaving FF intact (Lamme, Zipser, et al. 1998; Wollstadt et al. 2017), largely supresses hemodynamic measures such as BOLD and CBF(Qiu et al. 2008) while having little to no effect on spike rates(Pisauro et al. 2013). The latter is most likely due to the fact that FB can have both faciliatory and inhibitory effects on V1 (Nassi et al. 2013). In addition, the main determinant of a cortical neurons response to thalamic input is not in its strength, but rather in its synchrony(Alonso et al. 1996; Bruno and Sakmann 2006). Hemodynamic measures such as BOLD are relatively insensitive to changes in the latter(Butler, Bernier, et al. 2017a), again highlighting how fMRI measures in V1 are more likely shaped by the numerous FB rather than a few, albeit synchronized, FF inputs it receives. Taken together, our results provide direct experimental evidence that synaptic activity produced by FB input is clearly visible with fMRI but difficult to observe in the spiking output of cortical neurons (Logothetis 2008). This likely explains why classic neurophysiology results obtained in anesthetized preparations, such as contrast- invariant orientation tuning(Geisler and Albrecht 1997) and contrast-saturation, are not observed in human fMRI. As it stands, the source of this FB cannot be unambiguously delineated from the current study. At low-contrast, we found that alpha orientation profile closely matched the BOLD in both striate and extrastriate areas (V3, V4, LOC), indicative of both proximal and distal FB contributions. The latter may reflect that fact that such stimuli are more difficult to evaluate than high contrast stimuli and thus may encourage subjects to pay more attention(Lee and Maunsell 2010), thus leading to top-down attentional modulation of V1(Thiele and Bellgrove 2018). Assuming this was also the case for our subjects, our results suggest that attention (low-contrast stimuli) increases vertical BOLD responses, but decreases those at horizontal. This is in-line with the study by Liu et al. (Liu et al. 2007), who found that attending to a horizontal grating decreased the V1 BOLD response.

There have also been a few studies examining the relationship of the left atrium (LA) with CRT 13 14 . Until now, diastolic function, with the exception of atrio-ventricular dyssynchrony 15 , has not been expected to be reported when assessing a patient before CRT implantation. Nevertheless, the value of LA volume as a strong prognostic marker has largely been demonstrated in many fields, including systolic heart failure 16 . Furthermore, we can easily assess the size of the LA, as well as its function to some extent. Very promising observations have been made in the field of CRT 17 , including a study performed in our institution 14 , and even more observations have been made in the field of valvular heart disease

of the electromagnets linear domain, i.e. i 0 = i max /2 = 3.5A 4.3. Dynamic characterization and model identification One of the major issue to obtain FRF is to know the force level that is applied at a given frequency. Due to the low bandwidth frequency of the dynamometer table, this force level can’t be measured at high frequencies (≤ 1kHz). To evaluate the magnetic force from equations (8) and (9), the amplitude of flux density as function of the excitation

Competing Interests: The authors have declared that no competing interests exist. * E-mail: steven.treistman@upr.edu Introduction For many years, lipid perturbation was considered to be the primary molecular mechanism responsible for the actions of ethanol in the nervous system, responsible for downstream effects on protein function [1]. However, data collected more recently, especially using mutagenesis, have shifted the focus of the ethanol field to the direct actions of ethanol on membrane proteins [2,3,4]. However, work using highly reduced systems in which ethanol’s actions are analyzed in bilayers containing only a small number of lipids and a single species of membrane protein channel have made very clear that even after accepting the proposition that target proteins contain a binding site for ethanol, the lipids adjacent to the protein exert a profound influence on the response of the protein to the drug. Thus, any interpretation of drug action at the molecular level that considers the protein, but does not consider the contribution of its lipid environment is likely to be inadequate [5]. Among the attributes of lipids that influence ethanol’s actions on imbedded proteins are bilayer thickness,

At q  0 . 3 h Mpc − 1 , however, the measured response function is damped compared to the PT. The one-loop PT predicts the re- sponse function to reach a constant 10 ; at the two-loop order, it grows in amplitude with time. The numerical measurements show on the other hand that the scaled response function is strongly damped with decreasing redshift. It is such that the couplings take place effectively between modes of similar wavelengths. This effect is particularly important at late time. At redshift zero, the discrep- ancy between the model and simulations is striking. Furthermore, analysis of the response structure at three and higher loop order (see e.g., [14] ) suggests that PT calculations, at any finite order, predict an even larger amplitude of the response function in the high q region. This strongly suggests that this anomaly is genuinely non-perturbative.

Abstract Hemodynamic monitoring is necessary in unstable patients with shock, especially in the presence of complex circulatory compromise, high risk of rapid worsening or treat- ment intolerance. It allows early identification of the main mechanisms leading to shock states, hence helps in guiding adequate and targeted therapeutic interventions, and assesses both the efficacy and tolerance of therapy. Critical care echo- cardiography is currently recommended as the first-line tech- nique for the hemodynamic assessment of patients presenting with acute circulatory failure. Two complementary approa- ches may be used. Conventional transthoracic echocardiogra- phy is primarily performed and may be completed by the transesophageal approach in the presence of suboptimal ima- ging quality or if expected diagnostic accuracy is deemed insufficient. If a dedicated training in which modalities have recently been detailed is respected, critical care echocardio- graphy is ideally suited to best determine the type of shock and guide its therapeutic management. Importantly, critical care echocardiography provides additional information when compared with the standard management of patients with septic shock, and this accurately identifies the presence of acute cor pulmonale associated with the acute respiratory distress syndrome and depicts potential sources of impreci- sion of “blind” hemodynamic monitoring devices, such as the transpulmonary thermodilution. Repeated echocardiogra- phic assessment allows monitoring of both the efficacy an tolerance of therapeutic interventions, including the potential deleterious effects of ventilator settings on right ventricu- lar function in patients sustaining moderate-to-severe acute respiratory distress syndrome. In the near future, the emer- gence of miniaturized transesophageal echocardiographic probes promises to provide adequate tools for prolonged