Clinical

The computer model shows how the brain treats short-term memories

Scientists have developed a new theoretical model that shows how the brain uses particular types of neurons to retain short-term information. In a wide variety of neuropsychiatric conditions, including schizophrenia, as well as in normal aging, their findings may help shed light on why working memory is compromised. If you ever forget something only seconds after it was at the forefront of your mind—for example, the name of a dish you were about to order in a restaurant—then you know how important working memory is. This method of short-term memory is how, like the next move in a sequence of instructions, individuals retain knowledge for a matter of seconds or minutes to solve a problem or perform a task.

If you ever forget something only seconds after it was at the forefront of your mind—for instance, the name of a dish you were about to order at a restaurant—then you know how important working memory is. This type of short-term memory, like the next move in a sequence of instructions, is how individuals retain knowledge for a matter of seconds or minutes to solve a problem or perform a task. A new computational model has now been developed by Salk scientists to demonstrate how the brain uses particular types of neurons to retain short-term data. Their results, published in Nature Neuroscience on December 7, 2020, may help shed light on why in a wide variety of neuropsychiatric conditions, including schizophrenia, as well as in normal aging, working memory is damaged.

“Most research on working memory focuses on the excitatory neurons in the cortex, which are numerous and broadly connected, rather than the inhibitory neurons, which are locally connected and more diverse,” says Terrence Sejnowski, head of the Computational Neurobiology Laboratory of Salk and senior author of the new work. “However a recurrent neural network model that we taught to perform a working memory task surprised us by using inhibitory neurons to make correct decisions after a delay.”

A computer model of the prefrontal cortex, a region of the brain known to handle working memory, was developed in the new paper by Sejnowski and Robert Kim, a student of Salk and UC San Diego MD/PhD. To teach their model to carry out a test usually used to assess working memory in primates, the researchers used learning algorithms—the animals would decide if a pattern of colored squares on a screen matches one seen several seconds earlier. Sejnowski and Kim studied how their model was able to perform this task with high precision and then correlated it with current data on brain activity patterns shown in the success of the task by monkeys. The actual and simulated neurons involved in working memory worked on a slower timescale than other neurons in both experiments.

Kim and Sejnowski found that good working memory requires both the prevalence of long-time neurons and the close ties between inhibitory neurons, which suppress brain activity. The researchers could adjust how well the model performed on the working memory test as well as the timescale of the related neurons when they altered the frequency of interactions between these inhibitory neurons in their model.

The new findings point to the significance of inhibitory neurons and the researchers add, could encourage future research into the role of these cells in working memory. Studies on why certain persons with neuropsychiatric conditions, including schizophrenia and autism, struggle with working memory may also educate them.

Works Citation : Robert Kim, Terrence J. Sejnowski. Strong inhibitory signaling underlies stable temporal dynamics and working memory in spiking neural networks. Nature Neuroscience, 2020; DOI: 10.1038/s41593-020-00753-w

Categories: Clinical