Cortical computation in mammals and birds.
Cortical computation in mammals and birds.
复制标题
哺乳动物和鸟类的皮质计算。
DOI：
10.1073/pnas.1502209112
复制
发表时间：
2015
影响因子：
11.1
通讯作者：
Harris KD
中科院分区：
文献类型：
--
作者：
Harris KD
我们人类对我们的皮质特别自豪。考虑到我们的身体尺寸，我们的大脑比应有的要大。此外，我们的新皮质在大脑中所占的比例比所有其他哺乳动物都要大，而且我们的皮质可能比地球上任何其他物种含有更多的神经元 (1)。这种皮质扩张被认为使我们比动物王国的其他动物具有认知优势。然而，尽管我们的皮质可能更大，但它们的精细结构看起来与其他哺乳动物非常相似。人类皮层似乎包含相同的细胞类型，它们的接线和基因表达模式似乎与经过充分研究的模型系统（例如小鼠）基本相似。这一发现表明，随着哺乳动物的进化，常见的“规范皮质微电路”已被重新调整用途，以实现不同物种所需的不同类型的信息处理，包括在我们的例子中的语言和抽象推理 (2, 3)。在 PNAS 中，Calabrese 和 Woolley (4) 提供的数据表明，类似于哺乳动物皮层执行的计算也发生在鸟类中。人们通常认为哺乳动物比其他脊椎动物具有更先进的认知能力。皮层下结构被称为“爬行动物大脑”，并被认为只执行原始本能 (5)，而“鸟脑”仍然是校园里的侮辱。尽管如此，最近的研究至少消除了人们对鸟类可能非常聪明的怀疑。特别是，鸦科的成员能够完成曾经被认为是灵长类动物独有的认知任务：工具使用、令人印象深刻的时空记忆和明显的因果推理 (6)。在 YouTube 时代，鸟类智力的壮举正在病毒式传播：例如，日本腐肉乌鸦将坚果扔在人行横道上，让汽车从上面驶过，然后等待红灯变红且交通停止时收集坚果，这是一种非凡的方式（https://www.youtube.com/watch?v= BGPGknpq3e0）。鸟类没有新皮质；至少，它们没有与哺乳动物新皮质相匹配的六层大脑结构。然而，一个长期存在的理论认为，鸟类大脑皮层含有与哺乳动物新皮质同源的电路。尽管鸟类大脑皮层具有有核结构而不是层状结构，但传入连接模式和神经递质分布模式使卡滕推测鸟类大脑皮层中不同的细胞核与哺乳动物皮质的不同层是同源的 (7)。虽然这
We humans are particularly proud of our cortices. Our brains are bigger than they should be, given our body size; furthermore, our neocortices constitute a larger fraction of the brain than in all other mammals, and our cortices probably contain more neurons than those of any other species on the planet (1). This cortical expansion is thought to give us our cognitive edge over the rest of the animal kingdom. However, even though our cortices may be bigger, their fine structure appears quite similar to that of other mammals. The human cortex appears to contain the same cell types, and their patterns of wiring and gene expression appear basically similar to well-studied model systems, such as the mouse. This finding suggests that, as mammals evolved, a common “canonical cortical microcircuit” has been repurposed to implement the different types of information processing required by different species, including, in our case, language and abstract reasoning (2, 3). In PNAS, Calabrese and Woolley (4) present data that suggest that computations akin to those performed by the mammalian cortex occur also in birds. It is often assumed that mammals are more cognitively advanced than other vertebrates. Subcortical structures have been termed the “reptilian brain” and assumed to implement only primitive instincts (5), whereas “bird brain” remains a schoolyard insult. Nevertheless, recent research has removed any doubt that birds, at least, can be very smart. In particular, members of the corvid (crow) family are capable of cognitive tasks once thought exclusive to primates: tool use, impressive spatiotemporal memory, and apparent causal reasoning (6). In the age of YouTube, feats of bird intelligence are going viral: for example, the remarkable way Japanese Carrion Crows crack nuts by dropping them on a pedestrian crossing, letting cars drive over them, then waiting to collect the kernels once the lights turn red and the traffic stops (https://www. youtube. com/watch? v= BGPGknpq3e0). Birds don’t have a neocortex; at least, they don’t have a brain structure with six layers that match those of the mammalian neocortex. However, a long-standing theory holds that the avian pallium contains circuits homologous to those of the mammalian neocortex. Even though the avian pallium has a nucleated rather than laminated architecture, patterns of afferent connectivity and neurotransmitter distribution led Karten to hypothesize that distinct nuclei in the bird’s pallium are homologous to the different layers of the mammalian cortex (7). Although this