Dynamics of sensory processing: from neurons to behaviour

点击翻译按钮获取中文摘要

Imagine looking out the window of a high-speed train; nearby objects move quickly by, but a distant farmhouse appears relatively stationary. Now recall the different sensations from touching a rough surface with a motionless fingertip, and then with back-and-forth finger movements. In both cases, motion helps provide information about the environment, the distance of different objects in the visual example, and surface texture in the touch example. Besides being entirely different senses, the examples differ in another important way: the source of the motion. You passively observe the scene from the train, but you actively move your finger over the rough surface. An advantage of playing an active role is that you can move your finger at different speeds with different pressure to gather different information about the surface. This is referred to as “active sensing” and more generally means that energy is expended to acquire information about the environment. Self-motion is only one method of active sensing. Bats actively sense by producing ultrasound calls, a behaviour called echolocation. In active sensing, there is a “chicken-and-egg” problem. Sensory perception influences action, and action influences perception. For this reason, understanding how brain networks coordinate and control active sensing is a complicated task. My research focuses on an expert in active sensing, the weakly electric knifefish. These fish sense their environment using a self-generated oscillating electric field. Specialized electroreceptors on their skin encode modulations in this field produced by nearby objects. Sensing with electric fields enables these fish to navigate, capture prey and communicate in the dark, avoiding vision-dependent predators. This does not come without significant challenges. Electric field modulations are miniscule and often contaminated with high levels of background noise, including the electric signals of other fish – think of trying to follow a conversation of whispers in a noisy room. The fish overcomes these challenges by using two strategies. First, it actively controls its swimming movements – the “knifefish” namesake is a result of their stereotyped back-and-forth swimming movements (recall the finger movements discussed above used for active touch). Second, the oscillating electric field is extremely precise to allow even the smallest modulations to be detected. It is generated by a brain “pacemaker” network that is the least variable (most precise) of all known biological clocks. My proposal focuses on these strategies in the context of two fundamental questions: (1) How does motion influence perception? (2) How do brain networks control oscillatory activity? Our approach is multi-disciplinary by necessity. We combine the analysis of perception and behaviour, with single neuron electrophysiology and computational modeling, drawing from expertise in a range of fields, from neurophysiology and molecular biology to computer science and engineering. This multidisciplinary training ground is ideal for students at all levels, preparing them for a range of careers in biotechnology and high-technology, as well as in government labs and academia. A better understanding of brain oscillations and associated electric fields will impact diverse areas in neuroscience, from deep brain stimulation and neuroprosthetics to the etiology of epilepsy. Electric field-based sensing can be used in robotics, as well as new touch-screens. But these studies will also impact our understanding of sensing in general, with the hope that one day we will understand how brain processes underlie the memory of a high-speed train ride, the touch of a surface, or the electrosensory world of electric fish.