Detecting imminent collisions is critical for our survival. While some progress has recently been made on the related neural pathways in rodents, it remains unknown whether attention and consciousness is necessary for collision detection in human subcortical pathways. There are also spirited debates and significant unresolved issues on the neural substrates for the interesting and important phenomenon of "blindsight".

Using high-resolution 7T fMRI to study subcortical responses to looming stimuli in healthy participants and hemianopic patients, a research team led by Prof. ZHANG Peng and Prof. HE Sheng at the Institute of Biophysics, Chinese Academy of Sciences, discovered that subcortical pathways from the superior colliculus (SC) to the ventromedial pulvinar (vmPul) and ventral tegmental area (VTA) automatically detect collision trajectories even without attention and awareness to the visual stimuli. Behavioral performance, pupil reflexes and SC responses collectively indicated higher sensitivity to looming stimuli from the upper visual field, demonstrating an ecologically adaptive nature of the phylogenetically conserved tectal pathways. For hemianopic patients with unilateral lesions of the geniculostriate pathway, stronger response in the ipsilesional SC was associated with better detection performance of collision events in their blind visual field, concretely supporting a critical role of the tectofugal pathways in "blindsight" detection of threatening information.

These findings clearly demonstrate that human tectofugal pathways can automatically detect impending visual threats even without attention and awareness, and provide strong and direct evidence for them to serve as the neural substrates for "blindsight". These automatic subcortical mechanisms suggest an efficient and rapid neural computation of collision trajectory, which may inspire the optimization of computer vision algorithms for collision detection.

Figure. Subconscious detection of imminent collision in human subcortical pathways

Article link: https://doi.org/10.1371/journal.pbio.3002375

Contact: ZHANG Peng

Institute of Biophysics, Chinese Academy of Sciences

Beijing 100101, China

Email: zhangpeng@ibp.ac.cn

(Reported by Prof. ZHANG Peng's group)