Transcranial Magnetic Stimulation at M1 disrupts cognitive networks in schizophrenia

Uncertainty quantification of TMS simulations considering MRI segmentation errors

OBJECTIVE

Transcranial Magnetic Stimulation (TMS) is a non-invasive brain stimulation method that is used to study brain function and conduct neuropsychiatric therapy. Computational methods that are commonly used for electric field (E-field) dosimetry of TMS are limited in accuracy and precision because of possible geometric errors introduced in the generation of head models by segmenting medical images into tissue types. This paper studies E-field prediction fidelity as a function of segmentation accuracy.

APPROACH

The errors in the segmentation of medical images into tissue types are modeled as geometric uncertainty in the shape of the boundary between tissue types. For each tissue boundary realization, we then use an in-house boundary element method to perform a forward propagation analysis and quantify the impact of tissue boundary uncertainties on the induced cortical E-field.

MAIN RESULTS

Our results indicate that predictions of E-field induced in the brain are negligibly sensitive to segmentation errors in scalp, skull and white matter, compartments. In contrast, E-field predictions are highly sensitive to possible CSF segmentation errors. Specifically, the segmentation errors on the CSF and gray matter interface lead to higher E-field uncertainties in the gyral crowns, and the segmentation errors on CSF and white matter interface lead to higher uncertainties in the sulci. Furthermore, the uncertainty of the average cortical E-fields over a region exhibits lower uncertainty relative to point-wise estimates.

SIGNIFICANCE

The accuracy of current cortical E-field simulations is limited by the accuracy of CSF segmentation accuracy. Other quantities of interest like the average of the E-field over a cortical region could provide a dose quantity that is robust to possible segmentation errors.

Stimulation-Based Control of Dynamic Brain Networks

The ability to modulate brain states using targeted stimulation is increasingly being employed to treat neurological disorders and to enhance human performance. Despite the growing interest in brain stimulation as a form of neuromodulation, much remains unknown about the network-level impact of these focal perturbations. To study the system wide impact of regional stimulation, we employ a data-driven computational model of nonlinear brain dynamics to systematically explore the effects of targeted stimulation. Validating predictions from network control theory, we uncover the relationship between regional controllability and the focal versus global impact of stimulation, and we relate these findings to differences in the underlying network architecture. Finally, by mapping brain regions to cognitive systems, we observe that the default mode system imparts large global change despite being highly constrained by structural connectivity. This work forms an important step towards the development of personalized stimulation protocols for medical treatment or performance enhancement.

Brain stimulation is increasingly used in clinical settings to treat neurological disorders, but much remains unknown about how stimulation to a single brain region impacts large-scale, brain network activity. Using structural neuroimaging scans, we create computational models of brain dynamics for eight participants to explore how structure-function relationships constrain the effect of stimulation to a single region on the brain as a whole. Our results show that network control theory can be used to predict if the effects of stimulation remain focal or spread globally, and structural connectivity differentially constrains the effects of regional stimulation. Additionally, we study how stimulation of different cognitive systems spreads throughout the brain and find that stimulation of regions within the default mode network provide a mechanism to impart large change in overall brain dynamics through a densely connected structural network. By revealing how the stimulation of different brain regions and cognitive systems spreads differently through the brain, we provide a modeling framework to develop stimulation protocols to personalize medical treatments, enable performance enhancements, and facilitate cortical plasticity.

Transcranial magnetic stimulation (TMS) for schizophrenia

BACKGROUND

People with schizophrenia often experience symptoms which fail to fully respond to antipsychotic medication. Transcranial magnetic stimulation (TMS) has been proposed as a new treatment for people with schizophrenia, especially those who experience persistent auditory hallucinations.

OBJECTIVES

To estimate the effects of TMS alone, compared with sham TMS or with 'standard management' and any other comparison interventions in reducing psychotic symptoms associated with schizophrenia.

SEARCH METHODS

We searched the Cochrane Schizophrenia Group Trials Register (June 2006, June 2008, April 2013). This register is compiled by methodical searches of MEDLINE, EMBASE, BIOSIS, CINAHL, Dissertation abstracts, LILACS, PSYNDEX, PsycINFO, RUSSMED, and Sociofile, and is supplemented with handsearching of relevant journals and numerous conference proceedings.

SELECTION CRITERIA

We included all randomised controlled trials recruiting at least five participants and comparing TMS with sham TMS or any other treatment for people with schizophrenia.

DATA COLLECTION AND ANALYSIS

We extracted data independently. For dichotomous data we calculated relative risks (RRs) and their 95% confidence intervals (CIs). For continuous data, we calculated mean differences (MD) and 95% CI. We used a fixed-effect model. We assessed overall quality of the evidence using the GRADE approach.

MAIN RESULTS

We included 41 studies with 1473 participants in the review. We found significant differences in favour of temporoparietal TMS compared to sham TMS for global state measured on the CGI scale (7 RCTs, n = 224, MD -0.5, 95% CI -0.76 to -0.23, very low-quality evidence) and positive symptoms measured on the PANSS scale (5 RCTs, n = 127, MD -6.09, 95% CI -10.95 to -1.22, very low-quality evidence). Participants experienced significantly more headaches in the temporoparietal TMS group (10 RCTs, n = 392, RR 2.65, 95% CI 1.56 to 4.50, very low-quality evidence). However, no more participants left the study early from the TMS group than from the sham group (very low-quality evidence). Cognitive state was assessed using 39 different measures, and all were equivocal (very low-quality evidence).We included only two trials which compared temporoparietal TMS with standard treatment. In both trials the participants received first- and second-generation antipsychotic medication in both treatment groups, therefore TMS was used an adjunctive therapy to medication. We found no significant differences in the number of participants that showed clinical improvement in global state (1 RCT, n = 100, RR 1.19, 95% CI 0.91 to 1.57) or left the study early (2 RCTs, n = 140, RR 0.33, 95% CI 0.08 to 1.46) (both very low-quality evidence). No studies reported on global state score, mental state, cognitive state and adverse effects.For prefrontal TMS compared to sham TMS, global state was measured on three different scales, all of which presented equivocal results (very low quality evidence). We could not pool data for mental state on the PANSS scale due to high heterogeneity. Cognitive state was assessed using 19 different measures, with 15/19 being equivocal (very low-quality evidence). Prefrontal TMS caused more headaches (6 RCTs, n = 164, RR 2.77, 95% CI 1.22 to 6.26, very low-quality evidence) but there was no difference in the number of participants leaving the study early (very low-quality evidence). No studies reported data for clinical improvement.We found a significant difference in favour of prefrontal theta burst stimulation TMS compared to sham TMS for mental state on the PANNS scale (3 RCTs, n = 108, MD -5.71, 95% CI -9.32 to -2.10, very low evidence). We found no difference for clinical improvement, cognitive state, number of headaches, and leaving the study early (very low-quality evidence).None of the included studies reported satisfaction with care.

AUTHORS' CONCLUSIONS

Based on this review, there is insufficient evidence to support or refute the use of TMS to treat symptoms of schizophrenia. Although some evidence suggests that TMS, and in particular temporoparietal TMS, may improve certain symptoms (such as auditory hallucinations and positive symptoms of schizophrenia) compared to sham TMS, the results were not robust enough to be unequivocal across the assessment measures used. There was insufficient evidence to suggest any added benefit with TMS used as an adjunctive therapy to antipsychotic medication.The overall quality of evidence was graded as very low due to risk of bias, and this was accompanied by an imprecision in estimates due to the relatively small number of participants in the studies. Thus, consideration is required in improving the quality of trial processes, as well as the quality of reporting of ongoing and future TMS trials, so as to facilitate accurate future judgements in assessing risk of bias. Differences in TMS techniques in relation to stimulation intensity, stimulation length, brain areas stimulated and variations in the design of sham TMS all contributed to the heterogeneity of study findings and limited the interpretation and applicability of the results. In addition, the trials assessed their outcomes with a variety of scales, and usable data were limited. Therefore, to better evaluate the treatment effects of TMS in people with schizophrenia, we favour the use of standardised treatment protocols and outcome measures.

Uncertainty quantification in transcranial magnetic stimulation via high-dimensional model representation

A computational framework for uncertainty quantification in transcranial magnetic stimulation (TMS) is presented. The framework leverages high-dimensional model representations (HDMRs), which approximate observables (i.e., quantities of interest such as electric (E) fields induced inside targeted cortical regions) via series of iteratively constructed component functions involving only the most significant random variables (i.e., parameters that characterize the uncertainty in a TMS setup such as the position and orientation of TMS coils, as well as the size, shape, and conductivity of the head tissue). The component functions of HDMR expansions are approximated via a multielement probabilistic collocation (ME-PC) method. While approximating each component function, a quasi-static finite-difference simulator is used to compute observables at integration/collocation points dictated by the ME-PC method. The proposed framework requires far fewer simulations than traditional Monte Carlo methods for providing highly accurate statistical information (e.g., the mean and standard deviation) about the observables. The efficiency and accuracy of the proposed framework are demonstrated via its application to the statistical characterization of E-fields generated by TMS inside cortical regions of an MRI-derived realistic head model. Numerical results show that while uncertainties in tissue conductivities have negligible effects on TMS operation, variations in coil position/orientation and brain size significantly affect the induced E-fields. Our numerical results have several implications for the use of TMS during depression therapy: 1) uncertainty in the coil position and orientation may reduce the response rates of patients; 2) practitioners should favor targets on the crest of a gyrus to obtain maximal stimulation; and 3) an increasing scalp-to-cortex distance reduces the magnitude of E-fields on the surface and inside the cortex.

Numerical analysis and design of single-source multicoil TMS for deep and focused brain stimulation

Transcranial magnetic stimulation (TMS) is a tool for noninvasive stimulation of neuronal tissue used for research in cognitive neuroscience and to treat neurological disorders. Many TMS applications call for large electric fields to be sharply focused on regions that often lie deep inside the brain. Unfortunately, the fields generated by present-day TMS coils diffuse and decay rapidly as they penetrate into the head. As a result, they tend to stimulate relatively large regions of tissue near the brain surface. Earlier studies suggested that a focused TMS excitation can be attained using multiple nonuniformly fed coils in a multichannel array. We propose a systematic, genetic algorithm-based technique for synthesizing multichannel arrays that minimize the volume of the excited region required to achieve a prescribed penetration depth and maintain realistic values for the input driving currents. Because multichannel arrays are costly to build, we also propose a method to convert the multichannel arrays into single-channel ones while minimally materially deteriorating performance. Numerical results show that the new multi- and single-channel arrays stimulate tissue 2.4 cm into the head while exciting 3.0 and 2.6 times less volume than conventional Figure-8 coils, respectively.

Finding behavioral and network indicators of brain vulnerability

Resilience research has usually focused on identifying protective factors associated with specific stress conditions (e.g., war, trauma) or psychopathologies (e.g., post-traumatic stress disorder [PTSD]). Implicit in this research is the concept that resilience is a global construct, invariant to the unfavorable circumstances or the psychopathologies that may develop (i.e., the mechanisms underlying the resilience of an individual in all cases are expected to be similar). Here we contribute to the understanding of resilience—and its counterpart, vulnerability—by employing an approach that makes use of this invariant quality. We outline two main characteristics that we would expect from indicators of a vulnerable state: that they should appear across disorders regardless of specific circumstances, and that they should appear much before the disorder is evident. Next, we identify two sets of factors that exhibit this pattern of association with psychopathological states. The first was a set of “low-level” sensory, motor and regulatory irregularities that have been reported across the clinical literature; we suggest that these can serve as behavioral indicators of a vulnerable state. The second was the set of aberrations in network metrics that have been reported in the field of systems neuroscience; we suggest that these can serve as network indicators of a vulnerable state. Finally, we explore how behavioral indicators may be related to network indicators and discuss the clinical and research-related implications of our work.

Studies of properties of "Pain Networks" as predictors of targets of stimulation for treatment of pain

Two decades of functional imaging studies have demonstrated pain-related activations of primary somatic sensory cortex (S1), parasylvian cortical structures (PS), and medial frontal cortical structures (MF), which are often described as modules in a "pain network." The directionality and temporal dynamics of interactions between and within the cortical and thalamic modules are uncertain. We now describe our studies of these interactions based upon recordings of local field potentials (LFPs) carried out in an epilepsy monitoring unit over the one week period between the implantation and removal of cortical electrodes during the surgical treatment of epilepsy. These recordings have unprecedented clarity and resolution for the study of LFPs related to the experimental pain induced by cutaneous application of a Thulium YAG laser. We also used attention and distraction as behavioral probes to study the psychophysics and neuroscience of the cortical "pain network." In these studies, electrical activation of cortex was measured by event-related desynchronization (ERD), over SI, PS, and MF modules, and was more widespread and intense while attending to painful stimuli than while being distracted from them. This difference was particularly prominent over PS. In addition, greater perceived intensity of painful stimuli was associated with more widespread and intense ERD. Connectivity of these modules was then examined for dynamic causal interactions within and between modules by using the Granger causality (GRC). Prior to the laser stimuli, a task involving attention to the painful stimulus consistently increased the number of event-related causality (ERC) pairs both within the SI cortex, and from SI upon PS (SI > PS). After the laser stimulus, attention to a painful stimulus increased the number of ERC pairs from SI > PS, and SI > MF, and within the SI module. LFP at some electrode sites (critical sites) exerted ERC influences upon signals at multiple widespread electrodes, both in other cortical modules and within the module where the critical site was located. In summary, critical sites and SI modules may bind the cortical modules together into a "pain network," and disruption of that network by stimulation might be used to treat pain. These results in humans may be uniquely useful to design and optimize anatomically based pain therapies, such as stimulation of the S1 or critical sites through transcutaneous magnetic fields or implanted electrodes.

Attention to painful cutaneous laser stimuli evokes directed functional interactions between human sensory and modulatory pain-related cortical areas

The human 'pain network' includes cortical areas that are activated during the response to painful stimuli (termed category 1) or during psychological processes that modulate pain, for example, distraction (termed category 2). These categories include parts of the parasylvian (PS), medial frontal (MF), and paracentral cortex (S1&M1). Here we test the hypothesis that causal interactions both within and between category 1 and category 2 modules occur during attention to a painful stimulus. Event-related causality (ERC) was calculated from local field potentials recorded directly from these cortical areas during the response to a painful cutaneous laser stimulus in patients being monitored for epilepsy. The number of electrodes involved in pairs with significant ERC in category 1 was greater for pre-stimulus vs post-stimulus and for attention vs distraction. This is consistent with our prior evidence that the category 1 'pain network' changes rapidly with time intervals and tasks. In contrast, the interaction between categories was often unchanged or stable across intervals and tasks, particularly in MF. The proportion of contacts involved in interactions with PS was greater during distraction vs attention while activation was less, which suggests that distraction involves an inhibitory process in PS. Functional interactions between categories were overwhelmingly in the direction from category 2>1, particularly for contacts in MF which often had a driver role. These results demonstrate that MF is densely interconnected throughout the 'pain network' so that stimulation of MF might be used to disrupt the 'pain network' as a therapy for pain.

Fear conditioning is associated with dynamic directed functional interactions between and within the human amygdala, hippocampus, and frontal lobe

The current model of fear conditioning suggests that it is mediated through modules involving the amygdala (AMY), hippocampus (HIP), and frontal lobe (FL). We now test the hypothesis that habituation and acquisition stages of a fear conditioning protocol are characterized by different event-related causal interactions (ERCs) within and between these modules. The protocol used the painful cutaneous laser as the unconditioned stimulus and ERC was estimated by analysis of local field potentials recorded through electrodes implanted for investigation of epilepsy. During the prestimulus interval of the habituation stage FL>AMY ERC interactions were common. For comparison, in the poststimulus interval of the habituation stage, only a subdivision of the FL (dorsolateral prefrontal cortex, dlPFC) still exerted the FL>AMY ERC interaction (dlFC>AMY). For a further comparison, during the poststimulus interval of the acquisition stage, the dlPFC>AMY interaction persisted and an AMY>FL interaction appeared. In addition to these ERC interactions between modules, the results also show ERC interactions within modules. During the poststimulus interval, HIP>HIP ERC interactions were more common during acquisition, and deep hippocampal contacts exerted causal interactions on superficial contacts, possibly explained by connectivity between the perihippocampal gyrus and the HIP. During the prestimulus interval of the habituation stage, AMY>AMY ERC interactions were commonly found, while interactions between the deep and superficial AMY (indirect pathway) were independent of intervals and stages. These results suggest that the network subserving fear includes distributed or widespread modules, some of which are themselves "local networks." ERC interactions between and within modules can be either static or change dynamically across intervals or stages of fear conditioning.

[Production of rhythms in schizophrenia: the central role of attention]

BACKGROUND

Schizophrenia is a psychiatric illness that is characterised by a deficit in the fluent sequencing of thought and action. This problem of discoordination might be due to unreliable timing processes associated with a difficulty in allocating sufficient attention. In the present study, we placed ourselves within the hypothesis that schizophrenic patients may have difficulties in producing rhythmic tapping actions and that this deficit may be correlated with the degree of attention abnormalities.

METHOD

Subjects were required to tap in rhythm with alternating force levels and/or alternating time intervals (<1s) during trials lasting 24s. In addition, all patients performed an attention task (D2 test). A qualitative analysis of the tap trials was conducted in order to characterise the nature of the deficits that patients revealed.

RESULTS

Results showed that all patients revealed significant difficulties in performing the tapping trials. The number of trials removed was correlated with the level of attention dysfunction. Finally, our qualitative analysis revealed that 60% of patients presented attentional lapses - which were never observed in the healthy controls.

CONCLUSION

This study revealed deficits in the timing of action that resemble, at least on a behaviour level, the clinical lapses observed in schizophrenia. These lapses seem to be correlated to the degree of attention deficits. Future studies are now required in order to gain better understanding of the nature of the attention deficits in schizophrenia. More specifically, a better definition of the possible functional relationship between clinical lapses, cognitive lapses and action freezing is needed to develop innovating tools for rehabilitation.

A numerically optimized active shield for improved transcranial magnetic stimulation targeting

Transcranial magnetic stimulation (TMS) devices suffer of poor targeting and penetration depth. A new approach to designing TMS coils is introduced in order to improve the focus of the stimulation region through the use of actively shielded probes. Iterative optimization techniques were used to design different active shielding coils for TMS probes. The new approach aims to increase the amount of energy deposited in a thin cylindrical region below the probe relative to the energy deposited elsewhere in the region ("sharpness"), whereas, simultaneously increase the induced electric field deep in the target region relative to the surface ("penetration"). After convergence, the resulting designs showed that there is a clear tradeoff between sharpness and penetration that can be controlled by the choice of a tuning parameter. The resulting designs were tested on a realistic human head conductivity model, taking the contribution from surface charges into account. The design of choice reduced penetration depths by 16.7%. The activated surface area was reduced by 24.1% and the volume of the activation was reduced from 42.6% by the shield. Restoring the lost penetration could be achieved by increasing the total power to the coil by 16.3%, but in that case, the stimulated volume reduction was only 13.1% and there was a slight increase in the stimulated surface area (2.9%).

Transcranial magnetic stimulation

PURPOSE OF REVIEW

To present state-of-the-art transcranial magnetic stimulation (TMS) therapy, especially when it is used in psychiatric disorders, on the basis of an exhaustive literature search from 2006 to date (June 2008) on TMS papers published in Medline and Embase. Other references and comments from our own experience started 8 years ago have also been taken into account.

RECENT FINDINGS

The mechanism of action of TMS is now better understood. There is strong evidence of the safety and tolerability of TMS when standard protocols are used. The efficacy of the stimulation of the dorsolateral prefrontal cortex in depression is well documented, and there is evidence of the utility of TMS in posttraumatic stress disorder, in persistent auditory hallucinations in schizophrenia and in attention-deficit disorder with hyperactivity.

SUMMARY

There is enough evidence of the efficacy and safety of TMS in depression to include this technique in the therapeutic protocols of major depression. However, more research is needed on the use of this technique in other psychiatric and nonpsychiatric disorders such as posttraumatic stress disorder, persistent auditory hallucinations, attention-deficit disorder with hyperactivity and tinnitus.