Authors:

Vidal-Piñeiro, D., Sørensen, O., Strømstad, M., Amlien, I., Anderson, M., Baaré, W.F.C., Bartrés-Faz, D., Brandmaier, A.M., Bråthen, A.C., Garrido, P., Ghisletta P.; Grydeland, H., Henson, R.N., Rogier K.A., Kormacher, M., Kühn, S., Lindenberger, U., Mowinckel, A.M., Nyberg, L., Roe, J.M., Sneve, M.H., Sole-Padulles, C., Watne, L-O., Walhovd, K.B. & Fjell, A.M.

Reference:

Year of publication:

In Press

CBU number:

9156

Abstract:

In neuroimaging research, tracking individuals over time is key to understanding the interplay between brain changes and genetic, environmental, or cognitive factors across the lifespan. Yet, the extent to which we can estimate the individual trajectories of brain change over time with precision remains uncertain. In this study, we estimated the reliability of structural brain change in cognitively healthy adults from multiple samples and assessed the influence of follow-up time and number of observations. Estimates of cross-sectional measurement error and brain change variance were obtained using the longitudinal FreeSurfer processing stream. Our findings showed, on average, modest longitudinal reliability with 2 years of follow-up. Increasing the follow-up time was associated with a substantial increase in longitudinal reliability, while the impact of increasing the number of observations was comparatively minor. On average, 2-year follow-up studies require ≈2.7 and ≈4.0 times more individuals than designs with follow-ups of 4 and 6 years to achieve comparable statistical power. Subcortical volume exhibited higher longitudinal reliability than cortical area, thickness, and volume. The reliability estimates were comparable with those estimated from empirical data. The reliability estimates were affected by both the cohort’s age where younger adults had lower reliability of change and the preprocessing pipeline where the FreeSurfer’s longitudinal stream was notably superior than the cross-sectional stream. Suboptimal reliability inflated sample size requirements and compromised the ability to distinguish individual trajectories of brain aging. This study underscores the importance of long-term follow-ups and the need to consider reliability in longitudinal neuroimaging research.

URL:

Data for this project is available at: https://github.com/LCBC-UiO/Long_Brain_Reliability

Authors:

MONAGHAN, A., AKARCA, D., ASTLE, D.

Reference:

Year of publication:

2025

CBU number:

9155

Abstract:

What role do our genes play in shaping the structural organisation of the living human brain? Across a sample of 2,153 children (9–11 years old), we address this question, focusing on common genetic variants associated with cognitive ability and diffusion-based structural neuroimaging. Using polygenic scores, we test how variability in the genetic signal associated with cognitive ability is linked to simulated structural network properties, such as network efficiency. We fit a computational model to each connectome that simulates the emergence of high-level network properties. Central to the model is an economic trade-off between the “cost” of forming a given connection (a distance penalty) and the topological “value” that connection brings to the network. To simulate the network properties of those with the highest genetic propensity for cognitive ability, we had to use a significantly weaker wiring cost penalty. This softer distance penalty produces more stochastic, diverse, and efficient simulated networks. Further, those with a high genetic propensity for cognitive ability exhibited a more randomised simulated topology. Finally, we took a different approach to exploring the relationships between genes and model parameters by linking the distribution of those parameters with post-mortem gene expression data, with a comparative pathway enrichment analysis. Across the sample, overlapping biological and cellular pathways between polygenic scores and each child’s optimal cost-value trade-off emerged. Together, the generative wiring distance term, which varied maximally across participants but minimally across the cortex, was enriched for more ontologies than the wiring value term, which varied maximally across the cortex. However, the overlap in enriched ontologies between polygenic scores and the wiring value term was greater than that of polygenic scores and the wiring distance term. This application of computational modelling demonstrates that the underlying economic trade-offs needed to simulate the higher-order topological properties of networks vary according to genetic propensity for cognitive ability.

URL:

Data available, click to request

Authors:

GARCIA, C., CARLYON, R.P.

Reference:

Year of publication:

2025

CBU number:

9154

Abstract:

Objectives: Cochlear implant companies manufacture devices with different electrode array types. Some arrays have a straight geometry designed for minimal neuronal trauma, while others are precurved and designed to position the electrodes closer to the cochlear neurons. Due to their differing geometries, it is possible that the arrays are not only positioned differently inside the cochlea but also produce different patterns of the spread of current and of neural excitation. The panoramic electrically evoked compound action potential method (PECAP) provides detailed estimates of peripheral neural responsiveness and current spread for individual patients along the length of the cochlea. These estimates were assessed as a function of electrode position and array type, providing a normative dataset useful for identifying unusual patterns in individual patients. Design: ECAPs were collected from cochlear implant users using the forward-masking artifact-reduction technique for every combination of masker and probe electrode at the most comfortable level. Data were available for 91 ears using Cochlear devices, and 53 ears using Advanced Bionics devices. The Cochlear users had straight arrays (Slim Straight, CI-22 series, n = 35), or 1 of 2 precurved arrays (Contour Advance, CI-12 series, n = 43, or Slim Modiolar, CI-32 series, n = 13). Computed tomography scans were also available for 41 of them, and electrode-modiolus distances were calculated. The Advanced Bionics users had 1 of 2 straight arrays (1J, n = 9 or SlimJ, n = 20), or precurved arrays (Helix, n = 4 or Mid-Scala, n = 20). The ECAPs were submitted to the PECAP algorithm to estimate current spread and neural responsiveness along the length of the electrode array for each user. A linear mixed-effects model was used to determine whether there were statistically significant differences between different array types and/or for different electrodes, both for the PECAP estimates of current spread and neural responsiveness, as well as for the available electrode-modiolus distances. Correlations were also conducted between PECAP’s estimate of current spread and the electrode-modiolus distances. Results: For Cochlear users, significant effects of array type (p = 0.001) and of electrode (p < 0.001) were found on the PECAP’s current-spread estimate, as well as a significant interaction (p = 0.006). Slim Straight arrays had a wider overall current spread than both the precurved arrays (Contour Advance and Slim Modiolar). The interaction revealed the strongest effect at the apex. A significant correlation between PECAP’s current-spread estimate and the electrode-modiolus distances was also found across subjects (r = 0.516, p < 0.001). No effect of array type was found on PECAP’s estimate of current spread for the Advanced Bionics users (p = 0.979). Conclusions: These results suggest that for users of the Cochlear device, precurved electrode arrays show narrower current spread within the cochlea than those with lateral-wall electrode arrays, with the strongest effect present at the apex. No corresponding effects of array type were found in the Advanced Bionics device. This could have implications for device selection in clinical settings, although the authors underscore that this is a post-hoc analysis and does not demonstrate a causal link wherein device selection can be expected to give rise to specific neural excitation patterns.

URL:

Data available, click to request

Authors:

JACKSON, J.B., Scrivener, C.L., CORREIA, M., MADA, M., & Woolgar, A.

Reference:

Year of publication:

2025

CBU number:

9153

Abstract:

Background Transcranial magnetic stimulation (TMS) concurrent with functional magnetic resonance imaging (fMRI) can provide insights into the causal relationships between brain activity and behaviour. However, TMS pulses can cause artifacts in fMRI data, but these can be avoided if they are presented in short gaps between MRI slice acquisitions (interslice TMS-fMRI). New method We collected TMS-fMRI data to provide 1) guidance on the gap required and 2) a higher-level framework and code for researchers to test their own protocols. We quantified signal dropout and temporal signal-to-noise ratio in fMRI data (spherical phantom) for TMS pulses presented from up to 100ms before and after slice excitation. We delivered up to 3 pulses per volume with interslice gaps of 37.5ms/100ms (slice time 62.5ms), two 7-channel TMS-dedicated surface coils, and a multiband sequence (factor=2), on a Siemens 3T Prismafit scanner. We repeated a subset of parameters with a human participant. Results We observed minimal data contamination when pulses were applied at least -20ms/+50ms from slice excitation, and confirmed this approach can be used with 10Hz TMS. Comparison with existing methods Compared to other strategies that avoid TMS pulse-related artifacts, interslice allows for greater flexibility in terms of timing of the TMS pulse, MRI read out and any stimulus presentation. Conclusion A stimulation frequency faster than 10Hz would require a shorter gap or shorter slice acquisition times. Further, stimulator intensity, slice orientation, and the number of TMS pulses affected data quality and are important considerations for researchers when setting up their own protocol. Data on OSF (https://osf.io/tf5wj/

URL:

Data for this project is held by an external institution. Please contact the authors to request a copy.

Authors:

JAFARIAN, A., KARADAG ASSEM, M., KOCAGONCU, E., LANSKY, J.H., Fye, H., WILLIAMS, R., Quinn, A.J., Pitt, J., Raymont, V., Lowe, S., Singh, K.D., Woolrich, M., Nobre, A.C., HENSON, R.N., Friston, K.J., ROWE, J.B.

Reference:

Year of publication:

2025

CBU number:

9152

Abstract:

Neurodegenerative diseases, including Alzheimer's disease, are characterised by selective neuronal vulnerability with regional, laminar, cellular and neurotransmitter specificity. The regional losses of neurons and their synapses are associated with neurophysiological changes and cognitive decline. Hypotheses related to these mechanisms can be tested and compared by dynamic causal modelling (DCM) of human neuroimaging data, including magnetoencephalography (MEG). In this paper, we use DCM of cross-spectral densities to model changes between baseline and follow-up data in cortical regions of the default mode network, to characterise longitudinal changes in cortical microcircuits and their connectivity underlying resting-state MEG. Twenty-nine people with amyloid-positive mild cognitive impairment and Alzheimer's disease early dementia were studied at baseline and after an average interval of 16 months. To study longitudinal changes induced by Alzheimer's disease, we evaluate three complementary sets of DCM: (i) with regional specificity, of the contributions of neurons to measurements to accommodate regional variability in disease burden; (ii) with dual parameterisation of excitatory neurotransmission, motivated by preclinical and clinical evidence of distinct effects of disease on AMPA versus NMDA type glutamate receptors; and (iii) with constraints to test specific clinical hypothesis about the effects of disease-progression. Bayesian model selection at the group level confirmed evidence for regional specificity of the effects of Alzheimer's disease, with evidence for selective changes in NMDA neurotransmission, and progressive changes in connectivity within and between Precuneus and medial prefrontal cortex. Moreover, alterations in effective connectivity vary in accordance with individual differences in cognitive decline during follow-up. These applications of DCM enrich the mechanistic understanding of the pathophysiology of human Alzheimer's disease and inform experimental medicine studies of novel therapies. More generally, longitudinal DCM provides a potential platform for natural history and interventional studies of neurodegenerative and neuropsychiatric diseases, with selective neuronal vulnerability. The data that support the findings of this study can be requested from the senior author, noting that a data transfer agreement may be required under the terms of consent and data protection legislation. Anonymised data will also become available via https://www.dementiasplatform.uk/data-portal. The code associated with this manuscript is available at https://github.com/NIMG-22-2183/L-DCM_ HBM. The SPM12 software is openly available at https://www.fil.ion.ucl.ac.uk/spm/.

URL:

Authors:

WOOLGAR, A., Feredoes, E., ASSEM, M., Bassil, Y., Bergmann, T.O., Beynel, L., Burke, M., Cash, R.F.H., Comeau, R.M., CORREIA, M.M., Genc, E., Hartwigsen, G., Jackson, J.B., Kienle, M., Kunz, P., Leticevscaia, O., Luber, B., Lueckel, M., Mathiesen, C., Michael, E., Numssen, O., Oathes, D.J., Rosen, A.C., Schuhmann, T., Schuler, A.-L., Scrivener, C.L., Thielscher, A., Tik, M., Todorov, Y., Vasileiadi, M., Windischberger, C., Hermiller, M.S., and Sack, A.T.

Reference:

Year of publication:

2025

CBU number:

9151

Abstract:

Concurrent TMS-fMRI provides a step-change in the toolkit of neuroscience research. TMS enables the non-invasive perturbation of ongoing human brain activity and, coupled to fMRI for the simultaneous read-out of its effects across the brain, concurrent TMS-fMRI permits studies aiming at the causal inference of human brain-behaviour relationships, with implications for both fundamental research and clinical application. Many of the technical barriers to TMS-fMRI implementation, such as hardware design and setups, have now been solved and the clinical research community in the field is rapidly growing. Here, we discuss the results of an international consensus, from researchers at all levels and across the fields of cognitive and applied human neuroscience, on the experimental design and practical considerations of concurrent TMS-fMRI via 12 detailed use cases. These guidelines may facilitate the uptake of this approach and simplify the experimental design and planning stages.

URL:

Authors:

SHASHIDHARA, S., ASSEM, M., GLASSER, M.F., DUNCAN, J.

Reference:

Year of publication:

2024

CBU number:

9150

Abstract:

In the human brain, a multiple-demand (MD) network plays a key role in cognitive control, with core components in lateral frontal, dorsomedial frontal and lateral parietal cortex, and multivariate activity patterns that discriminate the contents of many cognitive activities. In prefrontal cortex of the behaving monkey, different cognitive operations are associated with very different patterns of neural activity, while details of a particular stimulus are encoded as small variations on these basic patterns (Sigala et al, 2008). Here, using the advanced fMRI methods of the Human Connectome Project and their 360-region cortical parcellation, we searched for a similar result in MD activation patterns. In each parcel, we compared multivertex patterns for every combination of three tasks (working memory, task-switching, and stop-signal) and two stimulus classes (faces and buildings). Though both task and stimulus category were discriminated in every cortical parcel, the strength of discrimination varied strongly across parcels. The different cognitive operations of the three tasks were strongly discriminated in MD regions. Stimulus categories, in contrast, were most strongly discriminated in a large region of primary and higher visual cortex, and intriguingly, in both parietal and frontal lobe regions adjacent to core MD regions. In the monkey, frontal neurons show a strong pattern of nonlinear mixed selectivity, with activity reflecting specific conjunctions of task events. In our data, however, there was limited evidence for mixed selectivity; throughout the brain, discriminations of task and stimulus combined largely linearly, with a small nonlinear component. In MD regions, human fMRI data recapitulate some but not all aspects of electrophysiological data from nonhuman primates.

URL:

Authors:

Moore, B.C.J., Humes, L.E., Cox, G., Lowe, D., Gockel, H.E.

Reference:

Year of publication:

2022

CBU number:

9149

Abstract:

Moore (2020) proposed a method for diagnosing noise-induced hearing loss (NIHL) sustained during military service, based on an analysis of the shapes of the audiograms of military personnel. The method, denoted M-NIHL, was estimated to have high sensitivity but low-to-moderate specificity. Here, a revised version of the method, denoted rM-NIHL, was developed that gave a better balance between sensitivity and specificity. A database of 285 audiograms of military noise-exposed men was created by merging two previously used databases with a new database, randomly shuffling, and then splitting into two, one for development of the revised method and one for evaluation. Two comparable databases of audiograms of 185 non-exposed men were also created, again one for development and one for evaluation. Based on the evaluation databases, the rM-NIHL method has slightly lower sensitivity than the M-NIHL method, but the specificity is markedly higher. The two methods have similar overall diagnostic performance. If an individual is classified as having NIHL based on a positive diagnosis for either ear, the rM-NIHL method has a sensitivity of 0.98 and a specificity of 0.63. Based on a positive diagnosis for both ears, the rM-NIHL method has a sensitivity of 0.76 and a specificity of 0.95.

URL:

Authors:

MONAGHAN, A., AKARCA, D., ASTLE, D.

Reference:

Year of publication:

2025

CBU number:

9148

Abstract:

What role do our genes play in shaping the structural organisation of the living human brain? Across a sample of 2,153 children (9–11 years old), we address this question, focusing on common genetic variants associated with cognitive ability and diffusion-based structural neuroimaging. Using polygenic scores, we test how variability in the genetic signal associated with cognitive ability is linked to simulated structural network properties, such as network efficiency. We fit a computational model to each connectome that simulates the emergence of high-level network properties. Central to the model is an economic trade-off between the “cost” of forming a given connection (a distance penalty) and the topological “value” that connection brings to the network. To simulate the network properties of those with the highest genetic propensity for cognitive ability, we had to use a significantly weaker wiring cost penalty. This softer distance penalty produces more stochastic, diverse, and efficient simulated networks. Further, those with a high genetic propensity for cognitive ability exhibited a more randomised simulated topology. Finally, we took a different approach to exploring the relationships between genes and model parameters by linking the distribution of those parameters with post-mortem gene expression data, with a comparative pathway enrichment analysis. Across the sample, overlapping biological and cellular pathways between polygenic scores and each child’s optimal cost-value trade-off emerged. Together, the generative wiring distance term, which varied maximally across participants but minimally across the cortex, was enriched for more ontologies than the wiring value term, which varied maximally across the cortex. However, the overlap in enriched ontologies between polygenic scores and the wiring value term was greater than that of polygenic scores and the wiring distance term. This application of computational modelling demonstrates that the underlying economic trade-offs needed to simulate the higher-order topological properties of networks vary according to genetic propensity for cognitive ability.

URL:

Authors:

DERMODY, N., Lorenz, R., Goddard, E., Villringer, A., WOOLGAR, A.

Reference:

Year of publication:

2025

CBU number:

9147

Abstract:

Attention enables the selective processing of relevant information. Two types of selective attention, spatial and feature-selective attention, have separable neural effects but in real life are often used together. Here, we asked how these types of attention interact to affect information coding in a frontoparietal ‘multiple-demand’ (MD) network, essential for attentional control. Using functional magnetic resonance imaging (fMRI) with multivariate pattern analysis, we examined how covert attention to object features (colour or shape) and spatial locations (left or right) influences coding of task-related stimulus information. We found that spatial and feature-selective attention interacted multiplicatively on information coding in MD and visual regions, such that there was above-chance decoding of the attended feature of the attended object and no detectable coding of visually equivalent but behaviourally irrelevant aspects of the visual display. The attended information had a multidimensional neural representation, with stimulus information (e.g., colour) and discrimination difficulty (distance from the categorical decision boundary) reflected in separate dimensions. Rather than boosting processing of whole objects or relevant features across space, our results suggest neural activity reflects precise tuning to relevant information, indicating a highly selective control process that codes behaviourally relevant information across multiple dimensions. 

URL:

Data available, click to request