Decision making   |   3D vision   |   Persistent activity

Decision making

A large body of work has suggested that two key stages of the primate dorsal stream perform well-understood functions: neurons in area MT code the direction and strength of a moving stimulus, and neurons in area LIP time-integrate the directional signals coming from MT– a direct neural instantiation of the accumulation of evidence. Recent work in our lab suggests that neurons in LIP carry a bevy of signals, some task-related, others not. We are currently applying advanced statistical analyses, novel multiple-neuron and multiple-area electrophysiological recordings, as well as reversible pharmacological inactivations to more precisely understand the flow of information from MT to LIP, and from other areas of the brain through LIP and related oculomotor structures.

Representative empirical publications:

Yates, J.L., Park, I.M., Katz, L.N., Pillow, J.P., & Huk, A.C. (2017). Functional dissection of signal and noise in MT and LIP during decision-makingNature Neuroscience. [advance link]

Katz, L.N.*, Yates, J.L.*, Pillow, J.W., & Huk, A.C. (2016). Dissociated functional significance of decision-related activity in the primate dorsal stream. Nature, 535: 285–288. [link]

Park, I.M., Meister, M.L.R., Huk, A.C., & Pillow, J. (2014). Encoding and decoding in parietal cortex during sensorimotor decision-making. Nature Neuroscience, 17 (10): 1395–1403.  [pdf]

A review:

Huk, A.C., Katz, L.N., Yates, J.L. (2017). The role of the lateral intraparietal area in (the study of) decision makingAnnual Review of Neuroscience, 40: 349-372. [link]


3D vision

We apply a variety of techniques (human psychophysics and fMRI, single-neuron recordings) to understand how the dynamic patterns of light falling upon the two retinae are decoded to extract the trajectory of moving objects through 3D. To date, this line of work has produced convergent evidence that motion processing stages typically thought to encode flat (2D) motion signals actually perform inter-ocular computations to represent 3D directions. It has also suggested that even simple sensory areas might be best thought of as carrying multiplexed information that can be differentially read out for a variety of uses.

Representative empirical publications:

Czuba, T.B., Huk, A.C., Cormack, L.K. & Kohn, A. (2014). Area MT encodes three-dimensional motion. Journal of Neuroscience, 34(47):15522–15533. [pdf]

Rokers, B., Cormack, L.K., & Huk, A.C. (2009). Disparity- and velocity- based signals for 3D motion perception in human MT+Nature Neuroscience, 12(8), 1050-1055. [pdf]

A review:

Huk, A.C., Czuba, T.B., Knoell, J., & Cormack, L.K. (2017). Binocular mechanisms of 3D motion processingAnnual Review of Vision Science[link] 


Persistent activity and memory

Persistent neural activity is classically thought of as the core computation that allows the brain to remember and contemplate information acquired from the senses. It is qualitatively ubiquitous in the primate brain, but exactly what the time course of persistent activity reflects, and how it is generated and maintained, are only starting to become tractable given new biological and analytic techniques. We are performing multi-area neural recordings, as well as microstimulation-record experiments, to dissect how primate oculomotor areas (like LIP and FEF) come to carry persistent activity, and how these signals relate to various forms of memory.

Representative empirical publications:

Meister, M.L.R., Hennig, J., & Huk, A.C. (2013). Signal multiplexing and single-neuron computations in macaque LIP during perceptual decision-makingJournal of Neuroscience, 33(6): 2254-2267. [pdf]