Combining task-evoked and spontaneous activity to improve pre-operative brain mapping with fMRI

Data-efficient resting-state functional magnetic resonance imaging brain mapping with deep learning

OBJECTIVE

Resting-state functional MRI (RS-fMRI) enables the mapping of function within the brain and is emerging as an efficient tool for the presurgical evaluation of eloquent cortex. Models capable of reliable and precise mapping of resting-state networks (RSNs) with a reduced scanning time would lead to improved patient comfort while reducing the cost per scan. The aims of the present study were to develop a deep 3D convolutional neural network (3DCNN) capable of voxel-wise mapping of language (LAN) and motor (MOT) RSNs with minimal quantities of RS-fMRI data.

METHODS

Imaging data were gathered from multiple ongoing studies at Washington University School of Medicine and other thoroughly characterized, publicly available data sets. All study participants (n = 2252 healthy adults) were cognitively screened and completed structural neuroimaging and RS-fMRI. Random permutations of RS-fMRI regions of interest were used to train a 3DCNN. After training, model inferences were compared using varying amounts of RS-fMRI data from the control data set as well as 5 patients with glioblastoma multiforme.

RESULTS

The trained model achieved 96% out-of-sample validation accuracy on data encompassing a large age range collected on multiple scanner types and varying sequence parameters. Testing on out-of-sample control data showed 97.9% similarity between results generated using either 50 or 200 RS-fMRI time points, corresponding to approximately 2.5 and 10 minutes, respectively (96.9% LAN, 96.3% MOT true-positive rate). In evaluating data from patients with brain tumors, the 3DCNN was able to accurately map LAN and MOT networks despite structural and functional alterations.

CONCLUSIONS

Functional maps produced by the 3DCNN can inform surgical planning in patients with brain tumors in a time-efficient manner. The authors present a highly efficient method for presurgical functional mapping and thus improved functional preservation in patients with brain tumors.

A method for reconstruction of interpretable brain networks from transient synchronization in resting-state BOLD fluctuations

Resting-state (rs) fMRI has been widely used to examine brain-wide large-scale spatiotemporal architectures, known as resting-state networks (RSNs). Recent studies have focused on the temporally evolving characteristics of RSNs, but it is unclear what temporal characteristics are reflected in the networks. To address this issue, we devised a novel method for voxel-based visualization of spatiotemporal characteristics of rs-fMRI with a time scale of tens of seconds. We first extracted clusters of dominant activity-patterns using a region-of-interest approach and then used these temporal patterns of the clusters to obtain voxel-based activation patterns related to the clusters. We found that activation patterns related to the clusters temporally evolved with a characteristic temporal structure and showed mutual temporal alternations over minutes. The voxel-based representation allowed the decoding of activation patterns of the clusters in rs-fMRI using a meta-analysis of functional activations. The activation patterns of the clusters were correlated with behavioral measures. Taken together, our analysis highlights a novel approach to examine brain activity dynamics during rest.

Functional network properties derived from wide-field calcium imaging differ with wakefulness and across cell type

To improve 'bench-to-bedside' translation, it is integral that knowledge flows bidirectionally-from animal models to humans, and vice versa. This requires common analytical frameworks, as well as open software and data sharing practices. We share a new pipeline (and test dataset) for the preprocessing of wide-field optical fluorescence imaging data-an emerging mode applicable in animal models-as well as results from a functional connectivity and graph theory analysis inspired by recent work in the human neuroimaging field. The approach is demonstrated using a dataset comprised of two test-cases: (1) data from animals imaged during awake and anesthetized conditions with excitatory neurons labeled, and (2) data from awake animals with different genetically encoded fluorescent labels that target either excitatory neurons or inhibitory interneuron subtypes. Both seed-based connectivity and graph theory measures (global efficiency, transitivity, modularity, and characteristic path-length) are shown to be useful in quantifying differences between wakefulness states and cell populations. Wakefulness state and cell type show widespread effects on canonical network connectivity with variable frequency band dependence. Differences between excitatory neurons and inhibitory interneurons are observed, with somatostatin expressing inhibitory interneurons emerging as notably dissimilar from parvalbumin and vasoactive polypeptide expressing cells. In sum, we demonstrate that our pipeline can be used to examine brain state and cell-type differences in mesoscale imaging data, aiding translational neuroscience efforts. In line with open science practices, we freely release the pipeline and data to encourage other efforts in the community.

Spatiotemporal structure of sensory-evoked and spontaneous activity revealed by mesoscale imaging in anesthetized and awake mice

Stimuli-evoked and spontaneous brain activity propagates across the cortex in diverse spatiotemporal patterns. Despite extensive studies, the relationship between spontaneous and evoked activity is poorly understood. We investigate this relationship by comparing the amplitude, speed, direction, and complexity of propagation trajectories of spontaneous and evoked activity elicited with visual, auditory, and tactile stimuli using mesoscale wide-field imaging in mice. For both spontaneous and evoked activity, the speed and direction of propagation is modulated by the amplitude. However, spontaneous activity has a higher complexity of the propagation trajectories. For low stimulus strengths, evoked activity amplitude and speed is similar to that of spontaneous activity but becomes dissimilar at higher stimulus strengths. These findings are consistent with observations that primary sensory areas receive widespread inputs from other cortical regions, and during rest, the cortex tends to reactivate traces of complex multisensory experiences that might have occurred in exhibition of different behaviors.

Clinical applications of magnetic resonance imaging based functional and structural connectivity

Advances in computational neuroimaging techniques have expanded the armamentarium of imaging tools available for clinical applications in clinical neuroscience. Non-invasive, in vivo brain MRI structural and functional network mapping has been used to identify therapeutic targets, define eloquent brain regions to preserve, and gain insight into pathological processes and treatments as well as prognostic biomarkers. These tools have the real potential to inform patient-specific treatment strategies. Nevertheless, a realistic appraisal of clinical utility is needed that balances the growing excitement and interest in the field with important limitations associated with these techniques. Quality of the raw data, minutiae of the processing methodology, and the statistical models applied can all impact on the results and their interpretation. A lack of standardization in data acquisition and processing has also resulted in issues with reproducibility. This limitation has had a direct impact on the reliability of these tools and ultimately, confidence in their clinical use. Advances in MRI technology and computational power as well as automation and standardization of processing methods, including machine learning approaches, may help address some of these issues and make these tools more reliable in clinical use. In this review, we will highlight the current clinical uses of MRI connectomics in the diagnosis and treatment of neurological disorders; balancing emerging applications and technologies with limitations of connectivity analytic approaches to present an encompassing and appropriate perspective.

The utilisation of resting-state fMRI as a pre-operative mapping tool in patients with brain tumours in comparison to task-based fMRI and intraoperative mapping: A systematic review

PURPOSE

Resting-state functional Magnetic Resonance Imaging (rs-fMRI) is suggested to be a viable option for pre-operative mapping for patients with brain tumours. However, it remains an open issue whether the tool is useful in the clinical setting compared to task-based fMRI (T-fMRI) and intraoperative mapping. Thus, a systematic review was conducted to investigate the usefulness of this technique.

METHODS

A systematic literature search of rs-fMRI methods applied as a pre-operative mapping tool was conducted using the PubMed/MEDLINE and Cochrane Library electronic databases following PRISMA guidelines.

RESULTS

Results demonstrated that 50% (six out of twelve) of the studies comparing rs-fMRI and T-fMRI showed good concordance for both language and sensorimotor networks. In comparison to intraoperative mapping, 86% (six out of seven) studies found a good agreement to rs-fMRI. Finally, 87% (twenty out of twenty-three) studies agreed that rs-fMRI is a suitable and useful pre-operative mapping tool.

CONCLUSIONS

rs-fMRI is a promising technique for pre-operative mapping in assessing the functional brain areas. However, the agreement between rs-fMRI with other techniques, including T-fMRI and intraoperative maps, is not yet optimal. Studies to ascertain and improve the sophistication in pre-processing of rs-fMRI imaging data are needed.

Deep Brain Stimulation Initiative: Toward Innovative Technology, New Disease Indications, and Approaches to Current and Future Clinical Challenges in Neuromodulation Therapy

Deep brain stimulation (DBS) is one of the most important clinical therapies for neurological disorders. DBS also has great potential to become a great tool for clinical neuroscience research. Recently, the National Engineering Laboratory for Neuromodulation at Tsinghua University held an international Deep Brain Stimulation Initiative workshop to discuss the cutting-edge technological achievements and clinical applications of DBS. We specifically addressed new clinical approaches and challenges in DBS for movement disorders (Parkinson's disease and dystonia), clinical application toward neurorehabilitation for stroke, and the progress and challenges toward DBS for neuropsychiatric disorders. This review highlighted key developments in (1) neuroimaging, with advancements in 3-Tesla magnetic resonance imaging DBS compatibility for exploration of brain network mechanisms; (2) novel DBS recording capabilities for uncovering disease pathophysiology; and (3) overcoming global healthcare burdens with online-based DBS programming technology for connecting patient communities. The successful event marks a milestone for global collaborative opportunities in clinical development of neuromodulation to treat major neurological disorders.

Reconstructing lost BOLD signal in individual participants using deep machine learning

Signal loss in blood oxygen level-dependent (BOLD) functional neuroimaging is common and can lead to misinterpretation of findings. Here, we reconstructed compromised fMRI signal using deep machine learning. We trained a model to learn principles governing BOLD activity in one dataset and reconstruct artificially compromised regions in an independent dataset, frame by frame. Intriguingly, BOLD time series extracted from reconstructed frames are correlated with the original time series, even though the frames do not independently carry any temporal information. Moreover, reconstructed functional connectivity maps exhibit good correspondence with the original connectivity maps, indicating that the model recovers functional relationships among brain regions. We replicated this result in two healthy datasets and in patients whose scans suffered signal loss due to intracortical electrodes. Critically, the reconstructions capture individual-specific information. Deep machine learning thus presents a unique opportunity to reconstruct compromised BOLD signal while capturing features of an individual's own functional brain organization.

Human cortical networking by probabilistic and frequency-specific coupling

Large-scale cortical networking patterns have been established based on the correlation of slow fluctuations of resting fMRI signals. However, the electrophysiological mechanism of cortical networking remained to be elucidated. With large-scale human ECoG recording, we developed a novel approach for functional network parcellation on the basis of probabilistic co-activation of cortical sites in spatio-temporal microstates. The parcellated networks were verified by electrical cortical stimulation (ECS) and somatosensory evoked potentials recording, which showed significantly higher accuracy than the traditional long-term correlation method. This provides direct electrophysiological evidence supporting the dynamic nature of cortical networking. Further analysis revealed that the brain-wide connectivity is likely established on the coupling of ECoG power envelop over a common carrier frequency ranging from alpha to low-beta (8-32Hz). Surprisingly, the cortical networking pattern over this specific frequency was found to be consistent across various tasks, which resembles the resting networks. The high similarity between the above functional network parcellation and the fMRI resting network atlas in individuals also suggested the slow power-envelope coupling of band-limited neural oscillations as the electrophysiological basis of spontaneous BOLD signals. Collectively, our findings on direct human recording revealed a probabilistic and frequency specific coupling mechanism for large-scale cortical networking shared by task and resting brain.

Limitations of functional neuroimaging for patient selection and surgical planning in glioma surgery

The optimal surgical management of gliomas requires a balance between surgical cytoreduction and preservation of neurological function. Preoperative functional neuroimaging, such as functional MRI (fMRI) and diffusion tensor imaging (DTI), has emerged as a possible tool to inform patient selection and surgical planning. However, evidence that preoperative fMRI or DTI improves extent of resection, limits neurological morbidity, and broadens surgical indications in classically eloquent areas is lacking. In this review, the authors describe facets of functional neuroimaging techniques that may limit their impact on neurosurgical oncology and critically evaluate the evidence supporting fMRI and DTI for patient selection and operative planning in glioma surgery. The authors also propose alternative applications for functional neuroimaging in the care of glioma patients.

Individualized Cortical Parcellation Based on Diffusion MRI Tractography

The spatial topological properties of cortical regions vary across individuals. Connectivity-based functional and anatomical cortical mapping in individuals will facilitate research on structure-function relationships. However, individual-specific cortical topographic properties derived from anatomical connectivity are less explored than those based on functional connectivity. We aimed to develop a novel individualized anatomical connectivity-based parcellation framework and investigate individual differences in spatial topographic features of cortical regions using diffusion magnetic resonance imaging (dMRI) tractography. Using a high-quality, repeated-session dMRI dataset (42 subjects, 2 sessions per subject), cortical parcels were derived through in vivo anatomical connectivity-based parcellation. These individual-specific parcels demonstrated good within-individual reproducibility and reflected interindividual differences in anatomical brain organization. Connectivity in these individual-specific parcels was significantly more homogeneous than that based on the group atlas. We found that the position, size, and topography of these anatomical parcels were highly variable across individuals and demonstrated nonredundant information about individual differences. Finally, we found that intersubject variability in anatomical connectivity was correlated with the diversity of anatomical connectivity patterns. Overall, we identified cortical parcels that show homogeneous anatomical connectivity patterns. These parcels displayed marked intersubject spatial variability, which may be used in future functional studies to reveal structure-function relationships in the human brain.

Performing group-level functional image analyses based on homologous functional regions mapped in individuals

Functional MRI (fMRI) studies have traditionally relied on intersubject normalization based on global brain morphology, which cannot establish proper functional correspondence between subjects due to substantial intersubject variability in functional organization. Here, we reliably identified a set of discrete, homologous functional regions in individuals to improve intersubject alignment of fMRI data. These functional regions demonstrated marked intersubject variability in size, position, and connectivity. We found that previously reported intersubject variability in functional connectivity maps could be partially explained by variability in size and position of the functional regions. Importantly, individual differences in network topography are associated with individual differences in task-evoked activations, suggesting that these individually specified regions may serve as the "localizer" to improve the alignment of task-fMRI data. We demonstrated that aligning task-fMRI data using the regions derived from resting state fMRI may lead to increased statistical power of task-fMRI analyses. In addition, resting state functional connectivity among these homologous regions is able to capture the idiosyncrasies of subjects and better predict fluid intelligence (gF) than connectivity measures derived from group-level brain atlases. Critically, we showed that not only the connectivity but also the size and position of functional regions are related to human behavior. Collectively, these findings suggest that identifying homologous functional regions across individuals can benefit a wide range of studies in the investigation of connectivity, task activation, and brain-behavior associations.

Use of an Individual-Level Approach to Identify Cortical Connectivity Biomarkers in Obsessive-Compulsive Disorder

BACKGROUND

Existing functional connectivity studies of obsessive-compulsive disorder (OCD) support a model of circuit dysfunction. However, these group-level observations have failed to yield neuroimaging biomarkers sufficient to serve as a test for the OCD diagnosis, predict current or future symptoms, or predict treatment response, perhaps because these studies failed to account for the substantial intersubject variability in structural and functional brain organization.

METHODS

We used functional regions, localized in each of 41 individual OCD patients, to identify cortical connectivity biomarkers of both global and dimension-specific symptom severity and to detect functional connections that track changes in symptom severity following intensive residential treatment.

RESULTS

Global OCD symptom severity was directly linked to dysconnectivity between large-scale intrinsic brain networks-particularly among the dorsal attention, default, and frontoparietal networks. Changes within a subset of connections among these networks were associated with symptom resolution. Additionally, distinct and nonoverlapping cortical connectivity biomarkers were identified that were significantly associated with the severity of contamination/washing and responsibility for harm/checking symptoms, highlighting the contribution of dissociable neural networks to specific OCD symptom dimensions. By contrast, when we defined functional regions conventionally, using a population-level brain atlas, we could no longer identify connectivity biomarkers of severity or improvement for any of the symptom dimensions.

CONCLUSIONS

Our findings would seem to encourage the use of individual-level approaches to connectivity analyses to better delineate the cortical and subcortical networks underlying symptom severity and improvement at the dimensional level in OCD patients.

Sensorimotor network parcellation for pre-surgical patients using low-pass filtered fMRI

Pre-surgical mapping of sensorimotor and language functions is crucial to reduce neurological deficits in epilepsy and tumor resection surgery. As non-invasive mapping, both resting-state and task-evoked functional MRI has been explored in pre-surgical mapping. In lack of standardized test paradigm, the reliability of fMRI mapping is still a concern for clinical use. In this study, to improve the reliability of fMRI based mapping, task fMRI data from all available task paradigms (motor movement, word repeating and picture naming) were low-pass filtered in the band of resting-state fMRI (0.01-0.08Hz) and concatenated to get more time points. With K-means clustering, it was shown that the sensorimotor network could be reliably parcellated into hand and tongue sub-regions. The resulted parcellations were further verified with invasive ECoG and ECS mapping. Both the accuracy and specificity were better than using the motor-task fMRI only. Especially, for those patients who failed in task fMRI mapping, our method was able to provide accurate mapping as well. Our results also indicate that cortical sensorimotor network pattern is intrinsic and always present during various tasks, which supports the physiological link between the spontaneous and the task-evoked BOLD signals.

Comparison between resting state fMRI networks and responsive cortical stimulations in glioma patients

OBJECTIVES

To validate the functional relevance of resting state networks (RSNs) by means of a comparison of resting state connectivity (RSC) between language regions elicited by direct cortical stimulation versus RSC between random regions; and to evaluate the accuracy of resting state fMRI in surgical planning by assessing the overlap between RSNs and intraoperative functional mapping results.

METHODS

Sensorimotor and language eloquent sites were identified by direct electrical cortical stimulation in 98 patients with a diffuse low-grade glioma. A seed to voxel analysis with inter-language stimulation point connectivity versus inter-random ROIs connectivity was performed (19 patients). An independant component analysis (ICA) was also applied to rsfMRI data. Language and sensorimotor components were selected over 20 independent components and compared to the corresponding stimulation points and resected cortex masks (31 and 90 patients, respectively).

RESULTS

Mean connectivity value between language seeds was significantly higher than the one between random seeds (0.68 ± 0.39 and 0.12 ± 0.21 respectively, P < 10^-10 ). 96 ± 11% of sensorimotor stimulation points were located within 10 mm from sensorimotor ICA maps versus 92 ± 21% for language. 3.1 and 15% of resected cortex overlapped sensorimotor and language networks, respectively. Mean sensorimotor stimulation points and resected cortex z-scores were 2.0 ± 1.2 and -0.050 ± 0.60, respectively (P < 10^-10 ). Mean language stimulation points and resected cortex z-scores were 1.6 ± 1.9 and 0.68 ± 0.91, respectively, P < 0.005.

CONCLUSION

The significantly higher RSC between language seeds than between random seeds validated the functional relevance of RSC. ICA partly succeeded to distinguish eloquent versus surgically removable areas and may be possibly used as a complementary tool to intraoperative mapping. Hum Brain Mapp 37:3721-3732, 2016. © 2016 Wiley Periodicals, Inc.

Combining Functional Studies with Intraoperative MRI in Glioma Surgery

Maximal safe resection is the cornerstone of treatment for low-grade and high-grade gliomas. In addition to high-resolution anatomic MRI studies that highlight tumor architecture, it is important to determine the relationship of the tumor to the eloquent cortical and subcortical areas to avoid introducing or exacerbating a neurologic deficit. The goal of this review was to highlight imaging modalities that provide functional information and can be integrated with intraoperative MRI navigation to maximize the extent of resection while preserving a patient's neurologic function.

iELVis: An open source MATLAB toolbox for localizing and visualizing human intracranial electrode data

BACKGROUND

Intracranial electrical recordings (iEEG) and brain stimulation (iEBS) are invaluable human neuroscience methodologies. However, the value of such data is often unrealized as many laboratories lack tools for localizing electrodes relative to anatomy. To remedy this, we have developed a MATLAB toolbox for intracranial electrode localization and visualization, iELVis. NEW METHOD: iELVis uses existing tools (BioImage Suite, FSL, and FreeSurfer) for preimplant magnetic resonance imaging (MRI) segmentation, neuroimaging coregistration, and manual identification of electrodes in postimplant neuroimaging. Subsequently, iELVis implements methods for correcting electrode locations for postimplant brain shift with millimeter-scale accuracy and provides interactive visualization on 3D surfaces or in 2D slices with optional functional neuroimaging overlays. iELVis also localizes electrodes relative to FreeSurfer-based atlases and can combine data across subjects via the FreeSurfer average brain.

RESULTS

It takes 30-60min of user time and 12-24h of computer time to localize and visualize electrodes from one brain. We demonstrate iELVis's functionality by showing that three methods for mapping primary hand somatosensory cortex (iEEG, iEBS, and functional MRI) provide highly concordant results. COMPARISON WITH EXISTING METHODS: iELVis is the first public software for electrode localization that corrects for brain shift, maps electrodes to an average brain, and supports neuroimaging overlays. Moreover, its interactive visualizations are powerful and its tutorial material is extensive.

CONCLUSIONS

iELVis promises to speed the progress and enhance the robustness of intracranial electrode research. The software and extensive tutorial materials are freely available as part of the EpiSurg software project: https://github.com/episurg/episurg.

Parcellating cortical functional networks in individuals

The capacity to identify the unique functional architecture of an individual's brain is a crucial step toward personalized medicine and understanding the neural basis of variation in human cognition and behavior. Here we developed a cortical parcellation approach to accurately map functional organization at the individual level using resting-state functional magnetic resonance imaging (fMRI). A population-based functional atlas and a map of inter-individual variability were employed to guide the iterative search for functional networks in individual subjects. Functional networks mapped by this approach were highly reproducible within subjects and effectively captured the variability across subjects, including individual differences in brain lateralization. The algorithm performed well across different subject populations and data types, including task fMRI data. The approach was then validated by invasive cortical stimulation mapping in surgical patients, suggesting potential for use in clinical applications.