Computation of object approach by a wide-field, motion-sensitive neuron

Area of Science:

• Neuroscience
• Computational Neuroscience
• Insect Vision

Background:

• The lobula giant motion detector (LGMD) in locusts is a key neuron for detecting approaching objects.
• LGMD's postsynaptic target, the descending contralateral motion detector (DCMD), also responds to these stimuli.

Purpose of the Study:

• Investigate the precise computation performed by the LGMD neuron in response to approaching objects.
• Determine the biophysical mechanisms underlying motion detection and potential multiplication operations in neurons.

Main Methods:

• Recorded postsynaptic activity of the DCMD neuron in locusts.
• Analyzed the relationship between object approach parameters and DCMD firing rate.
• Developed a computational model for LGMD/DCMD response based on angular threshold and edge velocity.

Main Results:

• Peak DCMD activity consistently occurred at a fixed angular threshold (thetathres: 15-40 degrees), regardless of object size or velocity.
• LGMD/DCMD's computation of thetathres was accurate, with errors corresponding to ommatidial separation.
• The firing rate was modeled as a product of angular edge velocity and an exponential function of angular size, suggesting a multiplication operation.
• The computation remained robust against changes in luminosity, contrast, and temperature.

Conclusions:

• The LGMD neuron implements a computation dependent on both angular size and edge velocity, likely involving a multiplication operation.
• LGMD serves as an excellent model for exploring the neural basis of multiplication in biological systems.
• This research provides insights into the computational strategies employed by insect visual systems for collision avoidance.