Few animals provide a readout that is as objective of their perceptual state as camouflaging cephalopods. Their skin display system includes an extensive array of pigment cells (chromatophores), each expandable by radial muscles controlled by motor neurons. If one could track the individual expansion states of the chromatophores, one would obtain a quantitative description—and potentially even a neural description by proxy—of the perceptual state of the animal in real time. Here we present the use of computational and analytical methods to achieve this in behaving animals, quantifying the states of tens of thousands of chromatophores at sixty frames per second, at single-cell resolution, and over weeks. We infer a statistical hierarchy of motor control, reveal an underlying low-dimensional structure to pattern dynamics and uncover rules that govern the development of skin patterns. This approach provides an objective description of complex perceptual behaviour, and a powerful means to uncover the organizational principles that underlie the function, dynamics and morphogenesis of neural systems.Tracking analyses of tens of thousands of individual chromatophores in freely behaving cephalopods enable studies of behaviour and development at cellular resolution.

中文翻译：

---

阐明墨鱼皮肤图案的控制和发展

很少有动物提供像伪装头足类动物一样客观的感知状态读数。他们的皮肤显示系统包括大量色素细胞（色素细胞），每个色素细胞都可以通过运动神经元控制的径向肌肉扩展。如果可以跟踪色素细胞的个体扩张状态，就可以实时获得动物感知状态的定量描述，甚至可能是通过代理进行的神经描述。在这里，我们展示了使用计算和分析方法在行为动物中实现这一目标，以每秒 60 帧、单细胞分辨率和数周时间量化数万个色素细胞的状态。我们推断出运动控制的统计层次结构，揭示模式动力学的潜在低维结构并揭示控制皮肤模式发展的规则。这种方法提供了对复杂感知行为的客观描述，以及揭示神经系统功能、动力学和形态发生背后的组织原理的有力手段。对自由行为的头足类动物中数万个个体色素细胞的跟踪分析能够研究行为和以细胞分辨率发展。