https://blue.feedland.org/?river=http://data.feedland.org/blue/feeds/DavidHaydock.xml Tue, 18 Feb 2025 12:52:38 GMT feedland v0.6.43 https://cyber.harvard.edu/rss/rss.html Tue, 18 Feb 2025 12:52:39 GMT http://rpc.rsscloud.io:5337/pleaseNotify Tue, February 18, 2025 7:52 AM EST <p>Ever heard of EEG microstates? They&#39;re usually defined as cluster centres of spatial topography, with parameters commonly used as biomarkers. Some study their transitional dynamics—&quot;microstate syntax&quot;. Our new review aims to discuss them: 🧵👇</p> <p>The commonality of syntax investigations is a discretisation of the EEG signal. Topographies at each time point are labelled with a microstate (whichever is the most similar to that time point). This makes a sequence of microstate symbols that are then subject to analysis.</p> <p>Studies on microstate syntax use a lot of different methods, and don’t always use the same process for defining the microstate sequence. Different terms are used for the same concepts, and documentation of the specifics of preprocessing and analysis steps can be lacking.</p> <p>In our new review, we organise existing methods into clear categories and define how different studies construct microstates, define microstate sequences, and how they go about investigating a sequence once they have it.</p> <p>A key issue highlighted (among others): is how microstate sequences are generated in and of themselves, in a “winner-takes-all” approach, where the complexity of the continuous EEG signal is simplified to a sequence of symbols.</p> <p>We argue that this common criticism can be investigated without throwing away microstate analysis by studying microstates in a continuous space (such as a t-SNE embedding, or similar). Doing so would allow for a richer understanding of microstate functional significance.</p> <p>Beyond this, we point out that existing methods which try to associate EEG microstates with fMRI patterns make sweeping assumptions when using GLM models by averaging the EEG time series to single values per TR, heavily simplifying the EEG signal.</p> <p>Our review provides a roadmap for researchers working on EEG microstate syntax. We hope to make results more comparable and useful for future studies, and call on researchers in the field to associate microstates to a continuous signal. </p> <p>If you work with EEG (and especially microstates) give the review a read! 🔗 <a href="https://doi.org/10.1016/j.neuroimage.2025.121090">https://doi.org/10.1016/j.neuroimage.2025.121090</a></p> Tue, 18 Feb 2025 12:52:38 GMT https://blue.feedland.org/?item=788790 https://blue.feedland.org/?item=788790 Ever heard of EEG microstates? They're usually defined as cluster centres of spatial topography, with parameters commonly used as biomarkers. Some study their transitional dynamics—"microstate syntax". Our new review aims to discuss them: 🧵👇 The commonality of syntax investigations is a discretisation of the EEG signal. Topographies at each time point are labelled with a microstate (whichever is the most similar to that time point). This makes a sequence of microstate symbols that are then subject to analysis. Studies on microstate syntax use a lot of different methods, and don’t always use the same process for defining the microstate sequence. Different terms are used for the same concepts, and documentation of the specifics of preprocessing and analysis steps can be lacking. In our new review, we organise existing methods into clear categories and define how different studies construct microstates, define microstate sequences, and how they go about investigating a sequence once they have it. A key issue highlighted (among others): is how microstate sequences are generated in and of themselves, in a “winner-takes-all” approach, where the complexity of the continuous EEG signal is simplified to a sequence of symbols. We argue that this common criticism can be investigated without throwing away microstate analysis by studying microstates in a continuous space (such as a t-SNE embedding, or similar). Doing so would allow for a richer understanding of microstate functional significance. Beyond this, we point out that existing methods which try to associate EEG microstates with fMRI patterns make sweeping assumptions when using GLM models by averaging the EEG time series to single values per TR, heavily simplifying the EEG signal. Our review provides a roadmap for researchers working on EEG microstate syntax. We hope to make results more comparable and useful for future studies, and call on researchers in the field to associate microstates to a continuous signal. If you work with EEG (and especially microstates) give the review a read! 🔗 https://doi.org/10.1016/j.neuroimage.2025.121090