Our brains have fairly precise records of when certain experiences or events occur, and this ability plays an important role in our daily lives. Scientists have found that the key to this lies in a structure in the brain called the hippocampus.
The hippocampus in the brain (red). | Image credit: Life Science Databases (LSDB) via Wiki Commons
The hippocampus in our brains gets its name because it is shaped like a real hippocampus, and it plays a key role in the temporal organization of memory and behavior. This includes not only our ability to remember when past experiences occurred, but also the ability to use this information from the past to imagine or predict future outcomes.
When this ability to remember is significantly impaired in neurological disorders, or is simply weakened with aging, our cognitive abilities are markedly problematic. But all along, we haven't developed a clear idea of the neuronal mechanisms behind these processes.
Recently, in a new study, scientists combined rodent electrophysiological recording technology with statistical machine learning analysis of large amounts of data to discover the basic mechanism by which the hippocampus sorts memory tissue, and how it can use this mechanism to plan future behavior. They believe that this finding may be a key step in understanding memory failure in cognitive impairment. The paper was recently published in Nature Communications.
Record the activity patterns of neurons
The research project took more than three years to complete and included experimental and data analysis phases. The researchers first trained the rats to perform a series of odor recognition tests and simultaneously monitored the firing of neurons in the rats' brains.
They released 5 different odors to the rats ,InSeq' and OutSeq, respectively, and rats needed to correctly identify whether each odor appeared sequentially through a behavioral response in the nose to receive a water reward.
Schematic of the experimental mechanism. Rats are rewarded by judging whether odors appear in order through nasal movements. | Image credit: Shahbaba, B. et al. (2022)
At the same time, the researchers monitored the activity patterns of the rat brains, looking at which cells were firing at any given moment and which were not, to detect exactly how the brain captured these sequential relationships.
During the minutes that the rats were tested, measurements of neuronal activity were taken at millisecond intervals, presenting a dynamic picture of how the brain was functioning. The team was able to "read" what the rats were thinking quickly and coherently by looking at the cell's "code," the pattern of cell discharge.
Advanced data analysis methods
Readings of hippocampal activity produce a large amount of raw data. Brain activity is recorded in milliseconds, and these experiments run for more than an hour, and the rate of data growth can be imagined. Neuroscientists need more advanced computing techniques to reach their goals. Therefore, from the beginning of the project, there are experts in data science involved.
When neurons encode information such as memories, scientists can peer into the process by examining all recorded neurons, the spike patterns of neuronal clusters. Data scientists realized in their analysis that they could apply these neural patterns as images to apply deep machine learning methods to them. The team chose convolutional neural networks to analyze the data, a method often used in image processing applications such as facial recognition.
Researchers use machine learning methods to analyze data. | Image source: UCI
In this way, the researchers were able to decode the firing patterns of neurons to retrieve information. They found that specific odor characteristics also reappear at different points in time, for example, when the rats in the experiment were thinking about the future and expecting what had not yet happened, some features were quickly replayed. These results prove that a fundamental function of the hippocampus network is precisely the order in which experiences are encoded, preserved, and predicted.
A case study of interdisciplinary collaboration
Scientists believe that the tools and methods developed in this study can be applied to a wide range of problems and extended to the study of other brain regions. This may be of great help to our understanding of Alzheimer's disease and other forms of dementia.
At the same time, they mentioned that this is an excellent example of interdisciplinary integration. Now, scientists in different fields have the opportunity to ask really important scientific questions that could not be studied in the past and to fundamentally think about the scientific method in a way that has never been seen before.
#创作团队:
Written by: Gaviota
Typography: Wenwen
#参考来源:
https://news.uci.edu/2022/02/15/uci-team-uncovers-key-brain-mechanisms-for-organizing-memories-in-time/
#图片来源:
Cover image: maxpixel
Header: maxpixel