Li Chuiwen, DePiero Victor J, Chen Hui, Tanabe Seiji, Cang Jianhua
Department of Psychology, University of Virginia, Charlottesville, VA 22904, USA.
Department of Biology, University of Virginia, Charlottesville, VA 22904, USA.
Curr Biol. 2025 Feb 24;35(4):723-733.e3. doi: 10.1016/j.cub.2024.12.029. Epub 2025 Jan 21.
Visual motion is a crucial cue for the brain to track objects and take appropriate actions, enabling effective interactions with the environment. Here, we study how the superior colliculus (SC) integrates motion information using asymmetric plaids composed of drifting gratings of different directions and speeds. With both in vivo electrophysiology and two-photon calcium imaging, we find that mouse SC neurons integrate motion direction by performing vector summation of the component gratings. The computation is constrained probabilistically by the possible physical motions consistent with each grating. Excitatory and inhibitory SC neurons respond similarly to the plaid stimuli. Finally, the probabilistically constrained vector summation also guides optokinetic eye movements. Such a computation is fundamentally different from that in the visual cortex, where motion integration follows the intersection of the constraints. Our studies thus demonstrate a novel neural computation in motion processing and raise intriguing questions regarding its neuronal implementation and functional significance.
视觉运动是大脑追踪物体并采取适当行动的关键线索，有助于与环境进行有效互动。在此，我们研究上丘（SC）如何利用由不同方向和速度的漂移光栅组成的不对称方格来整合运动信息。通过体内电生理学和双光子钙成像技术，我们发现小鼠SC神经元通过对组成光栅进行矢量求和来整合运动方向。这种计算在概率上受到与每个光栅一致的可能物理运动的限制。兴奋性和抑制性SC神经元对方格刺激的反应相似。最后，概率约束矢量求和也指导视动性眼球运动。这种计算与视觉皮层中的计算有根本不同，在视觉皮层中，运动整合遵循约束的交集。因此，我们的研究证明了运动处理中的一种新型神经计算，并提出了关于其神经元实现和功能意义的有趣问题。
