top of page

Recent Projects

Using Hidden Markov Modelling to explore variations in neural dynamics

We used machine learning to uncover the transient states of activity that characterise resting-state neural timeseries data in humans. We found that entropy predicts individual variations in general cognitive ability: people who have more chaotic patterns of switching between neural states perform better on cognitive tests. However, entropy isn't the whole story - flexibly switching into certain states, and out of other states, explains this relationship further.

To our knowledge, this was the first ever study to use HMMs in the analysis of neurophysiological data from a developmental sample. We're currently working on several follow-up studies, with very exciting results.

You can read our preprint here.

States.png
State properties plot.png
Transitions Cognition Relationships Plot.png

Using connectomics and graph theory to explore developmental heterogeneity in structural brain networks 

We used data from a large, heterogeneous sample of children with developmental difficulties to study how variations in the topological structure of brain networks relates to cognitive and behavioural traits. Using a combination of multidimensionality scaling and k-means clustering, we found that children with high levels of inattention and hyperactivity are best characterised into two 'neurotypes'. Children in these two clusters are comparable on over a dozen measures of behaviour, in addition to their age and gender distributions. 

However, the clusters differed in their general cognitive ability - one performed significantly worse on tests of visuospatial reasoning. Random walk simulations revealed that individuals from this cluster also had lower structural brain network communicability. Communicability, a measure of the efficiency with which information passes though a network, is emerging as a useful indicator of robust brain network development. 

So far, I have presented these findings at a number of conferences and Symposia, including Flux Congress 2022. You can read the paper here

Smaller clustering process figure.png
Graph Measures Schematic.png
Average connectome fig NEW.png
Nodal Comparison Diagram.png

Changing school practices for the better: an evidence-based approach

Inclusion is a mandated practice in UK schools, but one that suffers from a number of pitfalls, including striking inefficiencies in the provision of resources.

In 2022, we organised and facilitated a multi-disciplinary workshop as part of the Global Scientific Conference on Global Flourishing. This event brought together dozens of leaders in academic research, the charity sector, policy, education, and clinical practice to engage in collaborative processes aimed at improving school-level and national policies in the UK. More information about the event can be found here.

Following the workshop, in addition to feedback gathered at the It Takes All Kinds of Minds Conference (2023), we are producing a free, comprehensive set of evidence-based resources for the development of equitable school policies, as well as hosting a launch event for these materials that brought together relevant stakeholders. Following release, the resources will be available here. Feel free to sign up for our free launch event, happening online and at Robinson College, Cambridge on the 20th of September. 

BIS Logo.png
bottom of page