作者mulkcs (mulkcs)
看板Cognitive
標題[新知] 動物透過腦機介面來合作
時間Tue Jul 14 23:16:05 2015
這篇報導有兩篇文章, 但第二篇老鼠的研究我不是很懂就是了. 第一篇是用三隻猴子, 都
裝了腦機介面, 每隻猴子可以控制三維平面中的兩個維度. 最後三隻猴子一起可以控制三
個維度, 達成目標. 影片在原始網站:
http://0rz.tw/NJkdP
這研究對我來說比較值得關注的點是, 1)看來腦機介面已經可以很好的去解讀很簡單的運
動, 例如二維的活動, 否則誤差率大的話, 三隻猴子一起應該很難完成任務. 2) 不曉得猴
子是否知道他們的活動是互相合作或干擾情況下完成的, 或是只是想把點想辦法往要去的
地方控制. 如果是前者應該會很有趣, 表示可以有一個很好的動物模型來研究猴子經濟學
.看看猴子們的競合之類的.
--
Neuroscientists at Duke University have introduced a new paradigm for
brain-machine interfaces that investigates the physiological properties and
adaptability of brain circuits, and how the brains of two or more animals can
work together to complete simple tasks.
These brain networks, or Brainets, are described in two articles to be
published in the July 9, 2015, issue of Scientific Reports. In separate
experiments reported in the journal, the brains of monkeys and the brains of
rats are linked, allowing the animals to exchange sensory and motor
information in real time to control movement or complete computations.
In one example, scientists linked the brains of rhesus macaque monkeys, who
worked together to control the movements of the arm of a virtual avatar on a
digital display in front of them. Each animal controlled two of three
dimensions of movement for the same arm as they guided it together to touch a
moving target.
In the rodent experiment, scientists networked the brains of four rats
complete simple computational tasks involving pattern recognition, storage
and retrieval of sensory information, and even weather forecasting.
Brain-machine interfaces (BMIs) are computational systems that allow subjects
to use their brain signals to directly control the movements of artificial
devices, such as robotic arms, exoskeletons or virtual avatars.
The Duke researchers, working at the Center for Neuroengineering, have
previously built BMIs to capture and transmit the brain signals of individual
rats, monkeys, and even human subjects to artificial devices.
"This is the first demonstration of a shared brain-machine interface, a
paradigm that has been translated successfully over the past decades from
studies in animals all the way to clinical applications," said Miguel
Nicolelis, M.D., Ph. D., co-director of the Center for Neuroengineering at
the Duke University School of Medicine and principal investigator for the
study. "We foresee that shared BMIs will follow the same track, and could
soon be translated to clinical practice."
To complete the experiments, Nicolelis and his team outfitted the animals
with arrays implanted in their motor and somatosensory cortices to capture
and transmit their brain activity.
For one experiment highlighted in the primate article, researchers recorded
the electrical activity of more than 700 neurons from the brains of three
monkeys as they moved a virtual arm toward a target. In this experiment, each
monkey mentally controlled two out of three dimensions (i.e., x-axis and
y-axis; see video) of the virtual arm.
The monkeys could be successful only when at least two of them synchronized
their brains to produce continuous 3-D signals that moved the virtual arm. As
the animals gained more experience and training in the motor task,
researchers found that they adapted to the challenge.
The study described in the second paper used groups of three or four rats
whose brains were interconnected via microwire arrays in the somatosensory
cortex of the brain and received and transmitted information via those wires.
In one experiment, rats received temperature and barometric pressure
information and were able to combine information with the other rats to
predict an increased or decreased chance of rain. Under some conditions, the
authors observed that the rat Brainet could perform at the same level or
better than one rat on its own.
These results support the original claim of the same group that Brainets may
serve as test beds for the development of organic computers created by the
interfacing of multiple animal brains with computers.
Nicolelis and colleagues of the Walk Again Project, based in the project's
laboratory in Brazil, are currently working on a non-invasive human Brainet
to be used for neuro-rehabilitation training in paralyzed patients.
Explore further: Touch-sensing neurons are multitaskers
Journal:
http://medicalxpress.com/journals/scientific-reports/
網址:
http://medicalxpress.com/news/2015-07-neuroscientists-brain-to-brain-networks-
primates-rodents.html (
http://0rz.tw/NJkdP )
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 134.58.253.57
※ 文章網址: https://webptt.com/m.aspx?n=bbs/Cognitive/M.1436886968.A.CDE.html
另外一篇報導, 裡面描述當猴子合作一起解決問題時, 會學習的比較快. 也就是說, 通過
合作來控制, 問題會變簡單. 接下來他們會想測試使用困難的測驗下, 表現會不會比單一
猴子來的好. 這裡提出一個新概念, 或許可以把大腦當成運算元, 然後把大腦串聯起來變
成一個新的運算工具.
http://www.smithsonianmag.com/science-nature/linking-multiple-minds-could-help
-damaged-brains-heal-180955965/?no-ist (
http://tinyurl.com/qexlj7a )
※ 編輯: mulkcs (193.190.253.144), 07/21/2015 15:55:12