Methods of analyzing neural activity from calcium imaging

More and more researchers are using optical imaging to monitor activity in the brain. One type of optical imaging, two-photon calcium imaging, has gained widespread adoption due to its ability to record large neural populations. One key question, however, is what statistical methods should be used to analyze calcium imaging records.

“Before our work, two kinds of standards of Statistical methods has been applied to calcium Steve Chase, professor of biomedical engineering at Carnegie Mellon and the Institute of Neuroscience explains. either/or. Combining methods into one activity proved more successful.”

Researchers from Carnegie Mellon University and the Howard Hughes Medical Institute have teamed up to analyze existing methods being used to interpret calcium imaging records, as well as propose a new method that incorporates top two methods. The team’s new method, which performs deconvolution and dimensionality reduction simultaneously, is called the Calcium Imaging Linear Dynamic System, or CILDS.

“Through this process, we realized that current methods often focus on the activity of a single neuron, but since we know neurons carry and communicate important information about each other, it’s important to know that neurons are important to each other. we wanted to develop a method to test the neuron population to better summarize neural activity“, said Tze Hui Koh, first author of the paper and Carnegie Mellon’s biomedical engineering graduate student. “In the range of situations in which we tested CILDS, it outperformed other methods. other independent legislation.”

Understanding the different tradeoffs and options required to combine deconvolution and downscaling is another important lesson detailed in a recent paper. Natural Computational Science paper. CILDS represents something greater than the sum of its parts.

Byron Yu, professor of biomedical engineering, electrical and computer engineering concludes: “We have developed a tool and published a paper that shows it to be useful. “We have code that’s now publicly available, and our goal is for the broader field to do science and discover something about the brain with it.”