Researchers based in Iran and the US have developed a new method to estimate “effective brain connectivity”, which can be used to help understand the interactions between different brain areas. By feeding a computational model with real electroencephalogram (EEG) data, they were able to confirm previous findings of differences in brain connectivity patterns in autistic children compared to their typically developing peers.
Brain dynamics and connectivity play an important role in the diagnosis and analysis of various neurological disorders, such as autism. The influence that a certain brain area has over another, or in other words, the causal relationships between different brain areas, are described as “effective connectivity”. Impairments or abnormalities in these connections can lead to dysfunctional behavior and social problems.
A recent paper, published in the journal Biomedical Engineering, presents a new statistical method for estimating effective connectivity. In their study, the Iranian and American researchers applied the so-called “dual Kalman-based method” to EEG signals, which are a measure of electrical activity in the brain. The method is based on the Kalman filter – a mathematical algorithm that is generally used for modeling, estimation and prediction tasks.
The researchers found that the results of their method surpass the accuracy of others in the field. They were able to demonstrate this in the first part of their study by comparing real EEG signals with predictions produced by their model, called “estimated signals”. The estimation error (a measure of the accuracy of the model’s predictions) was in fact lower than that of other connectivity estimation methods. According to the researchers, another advantage of the dual Kalman-based method is that it does not require any predefined anatomical or physiological knowledge.
In the second phase of the study, the researchers investigated effective connectivity in autistic children and in control subjects. To do so, they took EEG readings of all of the children at rest before comparing the patterns of the separate groups’ brain signals and the connections between eight different brain regions. They found that autistic children showed decreased connectivity between active brain regions as well as an increased connectivity within regions.
According to the scientists, these results comply with previous research regarding the correlation between autism and abnormal brain connectivity patterns. For example, autistic children frequently struggle to recognize facial expressions; when the visual areas of the brain are not sufficiently connected to higher-order processing areas, the interpretation of visual cues is impaired, leading to problems discerning emotions.
Because the biological and functional basis of autism is still not fully understood, the researchers claim that brain connectivity analyses will be useful in gaining a better understanding of this multifaceted disorder.
Read the original article here:
Mehdi Rajabioun, Ali Motie Nasrabadi, Mohammad Bagher Shamsollahi & Robert Coben: Effective Brain connectivity estimation between active brain regions in autism with dual Kalman based method, 21.09.2019.