Sebastian Baecke

Predicting Decisions in Human Social Interactions Using Real-Time fMRI and Pattern Classification

Negotiation and trade typically require a mutual interaction while simultaneously resting in uncertainty which decision the partner ultimately will make at the end of the process. Assessing already during the negotiation in which direction ones counterpart tends would provide a tremendous advantage. Recently, neuroimaging techniques combined with multivariate pattern classification of the acquired data have made it possible to discriminate subjective states of mind on the basis of their neuronal activation signature. However, to enable an online-assessment of the participants mind state both approaches need to be extended to a real-time technique. By combining real-time functional magnetic resonance imaging (fMRI) and online pattern classification techniques, we show that it is possible to predict human behavior during social interaction before the interacting partner communicates a specific decision. Average accuracy reached approximately 70% when we predicted online the decisions of volunteers playing the ultimatum game, a well-known paradigm in economic game theory. Our results demonstrate the successful online analysis of complex emotional and cognitive states using real-time fMRI, which will enable a major breakthrough for social fMRI by providing information about mental states of partners already during the mutual interaction. Interestingly, an additional whole brain classification across subjects confirmed the online results: anterior insula, ventral striatum, and lateral orbitofrontal cortex, known to act in emotional self-regulation and reward processing for adjustment of behavior, appeared to be strong determinants of later overt behavior in the ultimatum game. Using whole brain classification we were also able to discriminate between brain processes related to subjective emotional and motivational states and brain processes related to the evaluation of objective financial incentives.

Building virtual reality fMRI paradigms: A framework for presenting immersive virtual environments

The advantage of using a virtual reality (VR) paradigm in fMRI is the possibility to interact with highly realistic environments. This extends the functions of standard fMRI paradigms, where the volunteer usually has a passive role, for example, watching a simple movie paradigm without any stimulus interactions. From that point of view the combined usage of VR and real-time fMRI offers great potential to identify underlying cognitive mechanisms such as spatial navigation, attention, semantic and episodic memory, as well as neurofeedback paradigms. However, the design and the implementation of a VR stimulus paradigm as well as the integration into an existing MR scanner framework are very complex processes. To support the modeling and usage of VR stimuli we developed and implemented a VR stimulus application based on C++. This software allows the fast and easy presentation of VR environments for fMRI studies without any additional expert knowledge. Furthermore, it provides for the reception of real-time data analysis values a bidirectional communication interface. In addition, the internal plugin interface enables users to extend the functionality of the software with custom programmed C++ plugins. The VR stimulus framework was tested in several performance tests and a spatial navigation study. According to the post-experimental interview, all subjects described immersive experiences and a high attentional load inside the artifical environment. Results from other VR spatial memory studies confirm the neuronal activation that was detected in parahippocampal areas, cuneus, and occipital regions.

Structural brain alterations in patients with lumbar disc herniation: a preliminary study.

Chronic pain is one of the most common health complaints in industrial nations. For example, chronic low back pain (cLBP) disables millions of people across the world and generates a tremendous economic burden. While previous studies provided evidence of widespread functional as well as structural brain alterations in chronic pain, little is known about cortical changes in patients suffering from lumbar disc herniation. We investigated morphometric alterations of the gray and white matter of the brain in patients suffering from LDH. The volumes of the gray and white matter of 12 LDH patients were determined in a prospective study and compared to the volumes of healthy controls to distinguish local differences. High-resolution MRI brain images of all participants were performed using a 3 Tesla MRI scanner. Voxel-based morphometry was used to investigate local differences in gray and white matter volume between patients suffering from LDH and healthy controls. LDH patients showed significantly reduced gray matter volume in the right anterolateral prefrontal cortex, the right temporal lobe, the left premotor cortex, the right caudate nucleus, and the right cerebellum as compared to healthy controls. Increased gray matter volume, however, was found in the right dorsal anterior cingulate cortex, the left precuneal cortex, the left fusiform gyrus, and the right brainstem. Additionally, small subcortical decreases of the white matter were found adjacent to the left prefrontal cortex, the right premotor cortex and in the anterior limb of the left internal capsule. We conclude that the lumbar disk herniation can lead to specific local alterations of the gray and white matter in the human brain. The investigation of LDH-induced brain alterations could provide further insight into the underlying nature of the chronification processes and could possibly identify prognostic factors that may improve the conservative as well as the operative treatment of the LDH.

Increasing the reliability of data analysis of functional magnetic resonance imaging by applying a new blockwise permutation method

A recent paper by Eklund et al. (2012) showed that up to 70% false positive results may occur when analyzing functional magnetic resonance imaging (fMRI) data using the statistical parametric mapping (SPM) software, which may mainly be caused by insufficient compensation for the temporal correlation between successive scans. Here, we show that a blockwise permutation method can be an effective alternative to the standard correction method for the correlated residuals in the general linear model, assuming an AR(1)-model as used in SPM for analyzing fMRI data. The blockwise permutation approach including a random shift developed by our group (Adolf et al., 2011) accounts for the temporal correlation structure of the data without having to provide a specific definition of the underlying autocorrelation model. 1465 publicly accessible resting-state data sets were re-analyzed, and the results were compared with those of Eklund et al. (2012). It was found that with the new permutation method the nominal familywise error rate for the detection of activated voxels could be maintained approximately under even the most critical conditions in which Eklund et al. found the largest deviations from the nominal error level. Thus, the method presented here can serve as a tool to ameliorate the quality and reliability of fMRI data analyses.

Ethanol modulates the neurovascular coupling

Despite some evidence of the underlying molecular mechanisms the neuronal basis of ethanol-induced effects on the neurovascular coupling that forms the BOLD (blood oxygenation level dependent) signal is poorly understood. In a recent fMRI (functional magnetic resonance imaging) study monitoring ethanol-induced changes of the BOLD signal a reduction of the amplitude and a prolongation of the BOLD signal were observed. However, the BOLD signal is assumed to consist of a complex superposition of different underlying signals. To gain insight how ethanol influences stimulus efficacy, oxygen extraction, transit time and vessel-related parameters the fMRI time series from the sensori-motor and the visual cortex were analyzed using the balloon model. The results show a region-dependent decrease of the stimulus efficacy to trigger a post-stimulus neurovascular response as well as a prolongation of the transit time through the venous compartment. Oxygen extraction, feedback mechanisms and other vessel-related parameters were not affected. The results may be interpreted as follows: the overall mechanisms of the neurovascular coupling are still acting well at the moderate ethanol level of about 0.8‰ (in particular the vessel-related parts), but the potency to evoke a neurovascular response is already compromised most obviously in the supplementary motor area responsible for complex synchronizing and planning processes.

A proof-of-principle study of multi-site real-time functional imaging at 3T and 7T: Implementation and validation

Real-time functional Magnetic Resonance Imaging (rtfMRI) is used mainly for neurofeedback or for brain-computer interfaces (BCI). But multi-site rtfMRI could in fact help in the application of new interactive paradigms such as the monitoring of mutual information flow or the controlling of objects in shared virtual environments. For that reason, a previously developed framework that provided an integrated control and data analysis of rtfMRI experiments was extended to enable multi-site rtfMRI. Important new components included a data exchange platform for analyzing the data of both MR scanners independently and/or jointly. Information related to brain activation can be displayed separately or in a shared view. However, a signal calibration procedure had to be developed and integrated in order to permit the connecting of sites that had different hardware and to account for different inter-individual brain activation levels. The framework was successfully validated in a proof-of-principle study with twelve volunteers. Thus the overall concept, the calibration of grossly differing signals, and BCI functionality on each site proved to work as required. To model interactions between brains in real-time, more complex rules utilizing mutual activation patterns could easily be implemented to allow for new kinds of social fMRI experiments.

Real-time classification of activated brain areas for fMRI-based human-brain-interfaces

Functional MR imaging (fMRI) enables to detect different activated brain areas according to the performed tasks. However, data are usually evaluated after the experiment, which prohibits intra-experiment optimization or more sophisticated applications such as biofeedback experiments. Using a human-brain-interface (HBI), subjects are able to communicate with external programs, e.g. to navigate through virtual scenes, or to experience and modify their own brain activation. These applications require the real-time analysis and classification of activated brain areas. Our paper presents first results of different strategies for real-time pattern analysis and classification realized within a flexible experiment control system that enables the volunteers to move through a 3D virtual scene in real-time using finger tapping tasks, and alternatively only thought-based tasks.

Changes in gray matter volume after microsurgical lumbar discectomy: a longitudinal analysis

People around the world suffer chronic lower back pain. Because spine imaging often does not explain the degree of perceived pain reported by patients, the role of the processing of nociceptor signals in the brain as the basis of pain perception is gaining increased attention. Modern neuroimaging techniques (including functional and morphometric methods) have produced results that suggest which brain areas may play a crucial role in the perception of acute and chronic pain. In this study, we examined 12 patients with chronic low back pain and sciatica, both resulting from lumbar disc herniation. Structural magnetic resonance imaging (MRI) of the brain was performed 1 day prior to and about 4 weeks after microsurgical lumbar discectomy. The subsequent MRI revealed an increase in gray matter volume in the basal ganglia but a decrease in volume in the hippocampus, which suggests the complexity of the network that involves movement, pain processing, and aspects of memory. Interestingly, volume changes in the hippocampus were significantly correlated to preoperative pain intensity but not to the duration of chronic pain. Mapping structural changes of the brain that result from lumbar disc herniation has the potential to enhance our understanding of the neuropathology of chronic low back pain and sciatica and therefore may help to optimize the decisions we make about conservative and surgical treatments in the future. The possibility of illuminating more of the details of central pain processing in lumbar disc herniation, as well as the accompanying personal and economic impact of pain relief worldwide, calls for future large-scale clinical studies.

Non-invasive high-resolution tracking of human neuronal pathways: diffusion tensor imaging at 7T with 1.2 mm isotropic voxel size

Diffusion tensor imaging (DTI) allows characterizing and exploiting diffusion anisotropy effects, thereby providing important details about tissue microstructure. A major application in neuroimaging is the so-called fiber tracking where neuronal connections between brain regions are determined non-invasively by DTI. Combining these neural pathways within the human brain with the localization of activated brain areas provided by functional MRI offers important information about functional connectivity of brain regions. However, DTI suffers from severe signal reduction due to the diffusion-weighting. Ultra-high field (UHF) magnetic resonance imaging (MRI) should therefore be advantageous to increase the intrinsic signal-to-noise ratio (SNR). This in turn enables to acquire high quality data with increased resolution, which is beneficial for tracking more complex fiber structures. However, UHF MRI imposes some difficulties mainly due to the larger B1 inhomogeneity compared to 3T MRI. We therefore optimized the parameters to perform DTI at a 7 Tesla whole body MR scanner equipped with a high performance gradient system and a 32-channel head receive coil. A Stesjkal Tanner spin-echo EPI sequence was used, to acquire 110 slices with an isotropic voxel-size of 1.2 mm covering the whole brain. 60 diffusion directions were scanned which allows calculating the principal direction components of the diffusion vector in each voxel. The results prove that DTI can be performed with high quality at UHF and that it is possible to explore the SNT benefit of the higher field strength. Combining UHF fMRI data with UHF DTI results will therefore be a major step towards better neuroimaging methods.

Mapping fine-scale anatomy of gray matter, white matter, and trigeminal-root region applying spherical deconvolution to high-resolution 7-T diffusion MRI

ObjectivesWe assessed the use of high-resolution ultra-high-field diffusion magnetic resonance imaging (dMRI) to determine neuronal fiber orientation density functions (fODFs) throughout the human brain, including gray matter (GM), white matter (WM), and small intertwined structures in the cerebellopontine region.Materials and methodsWe acquired 7-T whole-brain dMRI data of 23 volunteers with 1.4-mm isotropic resolution; fODFs were estimated using constrained spherical deconvolution.ResultsHigh-resolution fODFs enabled a detailed view of the intravoxel distributions of fiber populations in the whole brain. In the brainstem region, the fODF of the extra- and intrapontine parts of the trigeminus could be resolved. Intrapontine trigeminal fiber populations were crossed in a network-like fashion by fiber populations of the surrounding cerebellopontine tracts. In cortical GM, additional evidence was found that in parts of primary somatosensory cortex, fODFs seem to be oriented less perpendicular to the cortical surface than in GM of motor, premotor, and secondary somatosensory cortices.ConclusionWith 7-T MRI being introduced into clinical routine, high-resolution dMRI and derived measures such as fODFs can serve to characterize fine-scale anatomic structures as a prerequisite to detecting pathologies in GM and small or intertwined WM tracts.