Category Archives: Uncategorized

Darling, the kids are out RAM!

Working memory, the ability to hold things in mind and manipulate them, is very important for children and is closely linked to their success in school. For instance, limited working memory leads to difficulties with following instructions and paying attention in class (also see our previous post https://forgingconnectionsblog.wordpress.com/2015/02/05/adhd-and-low-working-memory-a-comparison/). A major research aim is to understand why some children’s working memory capacity is limited. All children start with lower working memory capacity that increases as they grow up.
We also know that working memory, like all mental functions, is supported by the brain. The brain undergoes considerable growth and reorganisation as children grow up. Most studies so far looked at brain structures that support working memory across development. However, some structure may be more important in younger children and some in older children.
Our new study investigates for the first time how the contribution of brain structures to working memory may change with age. For that, we tested a large number of children between 6 and 16 years on different working memory tasks. We looked at aspects of working memory concerned with storing information (locations, words) and manipulating it. The children also completed MRI scans to image their brain structure. We found that white matter connecting the two hemispheres, and white matter connecting occipital and temporal areas is more important for manipulating information held in mind in younger children, but less important in older ones. In contrast, the thickness of an area in the left posterior temporal lobe was more important in older kids. We think that these findings reflect increased specialisation of the working memory system as it develops from a distributed system in younger children that requires good wiring between different brain areas to a more localised system that is supported by high-quality local machinery. By analogy, imagine you were completing a work project. If you were collaborating with people, quality and speed are largely determined by how well the team communicates – this would be very difficult if you were trying to coordinate via mobile phones in an area with low reception. On the other hand, if one person completed the project, then the outcome would depend on the ability of this worker. The insights from this study will help us to better understand how working memory is constrained at different ages, which may allow us to design better inventions in the future to help children who struggle with working memory.

A preprint of this paper is available on BioArxiv: http://biorxiv.org/content/early/2016/08/15/069617

The analysis code is available on GitHub: https://github.com/joebathelt/WorkingMemory_and_BrainStructure_Code

OHBM 2016 – Impressions and Perspectives

I had the immense privilege to attend the annual meeting of the Organization for Human Brain Mapping (OHBM) in Geneva last week. OHBM is a fantastic venue to see the latest and greatest developments in the field of human neuroimaging and this year made no exception. The program was jam-packed with keynote lectures, symposia, poster presentations, educational courses, and many informal discussions. It is almost impossible to describe the full breadth of the meeting, but I will try to summarize the ideas and developments that were most interesting to me.

Three broad themes emerged for me: big data, methodological rigor, and new approaches to science. Big data was pervasive throughout the meeting with a large number of posters making use of huge databases and special interest talks focussing on the practicalities and promises of data sharing and meta-analysis. It seems like the field is reacting to the widely discussed reproducibility crises (http://www.apa.org/monitor/2015/10/share-reproducibility.aspx) and prominent review articles about the faults of the low sample sizes that so far have been common practice in neuroimaging (http://www.nature.com/nrn/journal/v14/n5/full/nrn3475.html). These efforts seem well suited to firmly establish many features of brain structure and function, especially around typical brain development. In the coming years, this is likely to influence publishing standards, education, and funding priorities on a wide scale. I hope that this will not lead to the infant being ejected with the proverbial lavational liquid. There is still a need to study small samples of rare populations that give a better insight into biological mechanisms, e.g. rare genetic disorders. Further, highly specific questions about cognitive and brain mechanisms that require custom assessments will probably continue to be assessed in smaller scale studies, before being rolled out for large samples.

A related issue that probably also arose from the replication discussions is methodological rigor. Symposia at OHBM2016 discussed many issues that had been raised in the literature, like the effect of head motion on structural and functional imaging, comparison of post-mortem anatomy with diffusion imaging, and procedures to move beyond statistical association. Efforts to move towards higher transparency of analysis strategies were also prominently discussed. This includes sharing of more complete statistical maps (more info here: http://nidm.nidash.org/specs/nidm-overview.html– soon to be available in SPM and FSL), tools for easier reporting and visualisation of analysis pipelines, and access to well described standard datasets. I can imagine a future in which analyses are published in an interactive format that allows for access to the data and the possibility to tweak parameters to assess the robustness of the results.

These exciting developments also pose some challenges. The trend towards large datasets also requires a new kind of analytic and theoretical approach. This leads to a clash between traditional scientific approach and big data science. Let me expand: The keynote lectures presented impressive work that was carried out in the traditional hypothesis-test-refine-hypothesis fashion. For instance, keynote speaker Nora Volkow, National Institute of Drug Abuse, presented a comprehensive account of dopamine receptors in human addiction based on a series of elegant, but conceptually simple PET experiments. In contrast to the traditional approach of collecting a few measurements to test a specific hypothesis, big data covers a lot of different measurements with a very broad aim. This creates the problem of high-dimensional data that need to be reduced to reach meaningful conclusions. Machine learning approaches emerged as a relatively new addition to the human neuroscience toolkit to tackle pervasive problems associated with this. There is great promise in these methods, but standards for reliability still need to be established and new theoretical developments are needed to integrate these findings with current knowledge. Hopefully, there will be closer communication between method developers and scientists applying these tools to human neuroscience as these methods mature.

A single-gene disorder affects brain organisation

Many children struggle with learning particular skills despite good access to learning opportunities. For instance, some children need more time and considerably more effort to learn language compared to their peers. In some cases, this is associated with other problems, e.g. in certain types of childhood epilepsy, in others, these difficulties occur in isolation, e.g. specific language impairment. It is vital to understand the processes that lead to these difficulties so that problems can be identified early and treated with the most effective interventions. Yet, understanding these developmental difficulties still poses a scientific challenge. We know that genetic predisposition plays a role based on heritability studies and intermediate difficulties in family members. However, learning difficulties are associated with a large number of genes that individually have only very small effects. A possible reason for this could be that developmental disorders that are defined on the basis of behaviour reflect a mixture of underlying biology. This heterogeneity makes it very difficult to establish any causative mechanism.

One way of getting around this conundrum is to study known genetic disorders that share some similarity with more common developmental syndromes. To this end, we investigated a case group of individuals with mutations in a particular gene. We established that this mutation is associated with disproportionate deficits in attention, language, and oro-motor control (Baker et al., 2015). In the current study, we explored the effect of this genetic mutation on the organisation of the brain network to understand how a genetic difference may lead to differences in thinking and behaviour. Brain regions with typically high expression of this gene showed the highest connectivity, which may indicate that it is important for the development of structural connections. Further, cases with mutations in this gene displayed reduced efficiency of information transfer in the brain network. These findings suggest that brain organisation may provide an important intermediate level of description that could help to reveal how genetic differences give rise to learning difficulties. We hope to extend this work to compare brain organisation between genetic groups and developmental disorders directly in the future.

 

The article about the study is now available as a preprint:  http://dx.doi.org/10.1101/057687

Interested readers can also retrieve the analysis scripts for this study here:  https://github.com/joebathelt/ZDHHC9_connectome

 

 

Reference:

Baker, K., Astle, D. E., Scerif, G., Barnes, J., Smith, J., Moffat, G., et al. (2015). Epilepsy, cognitive deficits and neuroanatomy in males with ZDHHC9 mutations. Annals of Clinical and Translational Neurology, n/a–n/a. http://doi.org/10.1002/acn3.196

The brain that determines itself – a story of ingrained native language

“Pardon, comment dit-on en Francais?” – I tried hard to learn French to keep up with the bilingual son of close friends, but failed miserably. DuoLingo and sub-titled TV gave me the edge early on, but even at the tender age of 2.5 years, he is already overtaking me. Of course, in my profession, this makes me wonder how the brain manages this remarkable feat of mastering several languages so rapidly at his age – and offer some explanatory solace for my failings.

A popular belief is that children’s brains are more plastic, which explains their greater ease of language learning. However, studies in developmental cognitive neuroscience suggest that reality is a bit more complicated than that. Research shows that the brain expects particular stimuli during important periods of plasticity. During these periods, the brain soaks up particular bits of information like a sponge. However, the brain also has to keep this plasticity check. Otherwise, it would be like building a tower with wet concrete. The plasticity of the wet concrete is useful at first to shape the structure. But one has to wait until is hardens to further build on it to avoid ending up with a pile of sludge. Similarly, once acquired, representations in the brain have to set and might be difficult to alter later on.

A recent study by Pierce and colleagues provides an excellent example of this type of research. They recruited three groups of school-age children to their study: a group of monolingual French children, a group of bilingual children who spoke Chinese and learned French before their third birthday, and a group of Chinese children who were adopted by French-speaking parents before their third birthday. Importantly, the last group consisted of children who were exposed to Chinese early on, but did not speak Chinese anymore after adoption. In the experiment, the children performed a test of verbal working memory. For this task, children were presented with a sequence of French pseudowords*. When prompted, the child had to recall the last pseudoword that he or she had just seen (0-back), the one before that (1-back), or the one two pseudowords before (2-back) while an MRI machine measured the oxygenation of blood in their brain.

Response times and accuracy were the same in all groups, but the brain responses showed interesting differences. Monolingual French children showed increased blood oxygenation in regions that are typically associated with verbal working memory tasks like this. In contrast, the bilingual Chinese group showed higher activation in areas involved in frontal and parietal areas involved in cognitive control. Crucially, the third group of children who were exposed to Chinese early, but only spoke French, displayed activations that were more similar to the bilingual group than the monolingual group. So, despite relatively early acquisition of the second language and only speaking the second language, the early exposure resulted in a similar brain response as in bilingual children.

Previous studies also found higher activation in executive areas in bilingual adults. It was thought that this arises because both languages are automatically activated. The bilingual constantly suppresses the language that is not needed in the current situation. This explanation was supported by higher performance of bilinguals on general tasks that require inhibition. However, these new developmental findings could mean that the recruitment of executive areas is not driven by bilingualism per se, but perhaps by differences in early development that persists even when only one language is used later in life.

Importantly, the study does not suggest that there is any advantage or disadvantage in being exposed to a different language early in life. After all, the performance of all children in terms of accuracy and reaction time was statistically indistinguishable. However, the results provide an interesting insight into cognitive development beyond the early-is-better mantra so frequently repeated in pop science publications.

 

Footnote:

* Pseudowords are pronouncable sequences of letters that resemble words that could be found in a dictionary of a language, but do not have any meaning. An example of a pseudoword in English is ‘shum’ or ‘dake’

Reference:

Pierce, L., Chen, J.-K., Delcenserie, A., Genesee, F. & Klein, D. (2015): “Past experience shapes ongoing neural patterns for language”. Nature Communications 6 http://www.nature.com/ncomms/2015/151201/ncomms10073/full/ncomms10073.html

Image credit:

Pieter Brueghel the Elder (1526/1530–1569): The Tower of Babel [retrieved from commons.wikimedia.org]

The developmental cognitive neuroscience of the Great British Bake-off

The Great British Bake-off has quickly become a fixture in the cultural life of the UK. For a few weeks the country is obsessed with the fluffiness of the ultimate Victoria sponge, the moistness of the ideal carrot cake, and the layering of the perfect Black Forest Gateau. For the poor souls who are not familiar with the format, a selection of amateur bakers is charged with making a certain type of baked good. Usually the contestants have some artistic licence to add interesting flavour combinations to the standard recipe. The creations are then judged on taste, presentation, texture and other aspects.

But how do we find the perfect cake and tell it apart from a mediocre one? The first line of judgment is pretty obvious. Cakes that are misshapen, under-baked, or horrible in taste are pretty easy to spot. However, the difference between good bakes is more subtle. Furthermore, there is a great variety of cakes and full marks in one category do not guarantee full marks in another. For instance, a sponge cake is supposed to be very light and fluffy, not too moist and with little taste on its own. In contrast, the carrot cake is much denser, contains more moisture, and is richly flavoured with spices. These categories are almost exclusive – one would not want a dense sponge cake or a flavourless carrot cake. It seems that bakers and cake connoisseurs posses a representation of cake categories that they can use to judge creations against something like the Platonic ideal of a certain cake type.

The formation of (cake) categories

Surely, even Mary Berry wasn’t born with a representation of the most angelic Angel cake and must have learned these categories along the way. Probably, the representation differentiated from simple categories like “sweet stuff” to a more refined representations liked “baked sweet stuff with a light texture and golden brown colour”. An important aspect of the formation of these representations is exposure to many different examples that enable the extraction of communalities that allow for category formation. This process is known as statistical learning. A classic example of statistical learning comes from studies about language acquisition in infants by Jenny Saffran and colleagues. One of the first steps in language learning for infants is to segment the continuous stream of speech into chunks and then to identify words from these chunks. For example, the infant would have to segment the speech stream ‘prettybaby’ into the constituent parts ‘pret-ty-ba-by’ and then identify their association with words, i.e. ‘pret-ty ba-by’ rather than ‘pre ty-ba by’. A possible way for the infant to achieve this feat is to learn which syllables frequently go together. In order to investigate this possibility, Saffran and colleagues exposed infants to different combinations of three syllables to create pseudowords that do not actually exist in the infants’ native language. After a short exposure, infants reacted differently to re-ordered combinations of the pseudowords compared to entirely new syllable combinations indicating that the infants had learned the statistical association between the syllables. Similarly, by exposure to many different examples of Sponge cake, we learn that they occupy a particular corner in the cake representational space in which features like golden colour, light texture, and buttery taste co-occur.

Neural correlates of category formation

Now, what are the underlying mechanisms in the brain that help us to achieve such a remarkable ability to extract communalities and form categories from the vast complexity of objects that we encounter? Unsurprisingly multiple brain areas need to interact to accomplish this task. On one hand, there are primary sensory regions that extract the lower level features of the stimulus in each modality, e.g. the basic shape and colour of the cake, the lemony smell etc. Many of these features are shared between different types of cake and are not specific to a cake category. However, these features can be connected to previous encounters in memory. This is mediated by a medial temporal system, particularly circuits of the hippocampus. Particular combinations of features may evoke a memory that links to a particular type of cake, e.g. the memory of the marvellous madeleines that Grandma Proust used to make.

However, category learning extends beyond retrieval of explicit memories. Freedman and colleagues identified a neural correlate of this categorisation in neurons in the prefrontal cortex. They trained monkeys to categories pictures of cats and dogs that were morphed to lie along a continuum. Neurons in the prefrontal cortex corresponded to their categorisation, e.g. “on” for picture categorised as a cat and “off” for a picture categorised as a dog, even though the stimuli themselves were continuous. These sharp categorisations may arise through feedback loops between the prefrontal cortex and the mesolimbic system involved in reward and reward prediction. Through the integration of these systems, learning about the association between the a particular stimulus and a reward is generalised to other similar stimuli.

As these examples show, the art of baking links to the deepest mysteries of how mind and brain make sense of a complex world based on limited experience and from a very early age. So, enjoy your next bake – it might expand the palette of your categorical representation!

References

Saffran, J. R., Aslin, R. N., & Newport, E. L. (1996). Statistical Learning by 8-Month-Old Infants. Science (New York, N.Y.), 274(5294), 1926–1928. http://doi.org/10.1126/science.274.5294.1926

Seger, C. A., & Miller, E. K. (2010). Category learning in the brain. Annual Review of Neuroscience, 33(1), 203–219. http://doi.org/10.1146/annurev.neuro.051508.135546