[新知] Hacking Memory

看板Cognitive作者mulkcs (mulkcs)時間10年前 (2015/03/10 07:49)推噓3(3推 0噓 6→)

留言9則, 5人參與討論串1/1

How to Insert a Memory Into the Brain of a Sleeping Mouse The lab mice awoke with happy memories… that researchers had inserted into their brains while they slept. New research in Nature Neurscience is the latest proof that we may soon live in a Philip K. Dick short story, where synthetic memories can be created via neural stimulation. Hyperbole aside, here’s how they did it. The researchers set out to test the following hypothesis: Animals consolidate memories while sleeping by reactivating neurons associated with the remembered experience. In five mice, the researchers used a clearly defined spatial memory. Each mouse had electrodes implanted in its hippocampus, the structure associated with memory. The electrodes recorded neural activity while the mouse explored a new environment. By monitoring the recorded signals, the researchers could identify spikes of electrical activity in certain neurons that were associated with a certain place in the chamber. Then each mouse took a one-hour nap. During that snooze the researchers continued to watched the signals from the hippocampus. They waited for moments when those place-associated neurons lit up with activity, suggesting that the mouse was recalling its experience. (Interestingly, this didn’t occur often during the REM sleep in which we dream.) On that activity cue, the researchers used a second set of electrodes to stimulate the brain’s medial forebrain bundle, a structure associated with the sensation of pleasure. Essentially, the scientists were teaching the mouse to associate that certain location with a reward. When each mouse awoke, it was sent back into the chamber, while the researchers watched to see where it would hang out. These five mice did not wander at random, but instead showed a clear preference for the place now associated with good feelings. It’s a small study, but a fascinating contribution to the new field of memory hacking. Another recent experiment showed a way of turning bad memories into good ones in mice, and DARPA is funding research on memory prosthetics for humans. One day we may look back on these early experiments, and remember them fondly—thanks to the electrodes implanted in our brains. 網址: http://spectrum.ieee.org/tech-talk/biomedical/bionics/how-to-insert-a-memory-in to-the-brain-of-a-sleeping-mouse ( http://0rz.tw/QPEh6 ) 論文 http://www.nature.com/neuro/journal/vaop/ncurrent/full/nn.3970.html#affil-auth 大意是說，先在老鼠的Hippocampus植入電極。讓老鼠醒著時，去熟悉新的環境。並且在此過程中，都會紀錄Hippocampus在各個新環境神經電訊號。然後讓老鼠小睡一下。在睡眠過程中，發現Hippocampus有類似反應時，表示Hippocampus正在recalling (我覺得recall不是很的辭彙, retrieve或consolidate會更好)，在此同時，對老鼠的medial frontal刺激達到reward的效果。這樣老鼠醒來時，在會比較偏好在有被電刺激medial frontal的新環境。我的直覺告訴我，這應該是某些實驗過程中的意外發現。畢竟電medial frontal對我來說效果太複雜。不是說電了就有reward就可以解釋全部。對於medial frontal和hippocampus 來說，我自己是覺得還蠻神奇的。 -- ※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 193.190.253.144 ※ 文章網址: https://www.ptt.cc/bbs/Cognitive/M.1425944970.A.6F6.html

推

ywwang

03/17 02:46, , 1^F

03/17 02:46, 1^F

推

bisconect

03/22 18:01, , 2^F

03/22 18:01, 2^F

→

bisconect

03/22 18:02, , 3^F

03/22 18:02, 3^F