Fast Recurrent Processing via Ventrolateral Prefrontal Cortex Is Needed by the Primate Ventral Stream for Robust Core Visual Object Recognition

• Published in: Neuron (Cambridge, Mass.) Mass.), 2021-01, Vol.109 (1), p.164-176.e5
• Main Authors: Kar, Kohitij, DiCarlo, James J.
• Format: Article
• Language: English
• Subjects: Animals; Behavior; core object recognition; Cortex (temporal); deep neural networks; Eye Movements - physiology; Firing pattern; inferior temporal cortex; Macaca mulatta; Male; Monkeys & apes; muscimol; Pattern recognition; Pattern Recognition, Visual - physiology; Photic Stimulation - methods; Population; population codes; Prefrontal cortex; Prefrontal Cortex - diagnostic imaging; Prefrontal Cortex - physiology; ventrolateral PFC; Visual discrimination; Visual pathways; Visual Perception - physiology
• ISSN: 0896-6273; 1097-4199
• DOI: 10.1016/j.neuron.2020.09.035

Distributed neural population spiking patterns in macaque inferior temporal (IT) cortex that support core object recognition require additional time to develop for specific, “late-solved” images. This suggests the necessity of recurrent processing in these computations. Which brain circuits are responsible for computing and transmitting these putative recurrent signals to IT? To test whether the ventrolateral prefrontal cortex (vlPFC) is a critical recurrent node in this system, here, we pharmacologically inactivated parts of vlPFC and simultaneously measured IT activity while monkeys performed object discrimination tasks. vlPFC inactivation deteriorated the quality of late-phase (>150 ms from image onset) IT population code and produced commensurate behavioral deficits for late-solved images. Finally, silencing vlPFC caused the monkeys’ IT activity and behavior to become more like those produced by feedforward-only ventral stream models. Together with prior work, these results implicate fast recurrent processing through vlPFC as critical to producing behaviorally sufficient object representations in IT. •Reversible inactivation of vlPFC induced specific deficits in object recognition•Induced IT population decode deficits were specific to “late-solved” images•Deficits in object recognition behavior were higher for late-solved images•vlPFC inactivation causes IT responses to better match feedforward models Kar and DiCarlo show that reversibly inactivating parts of macaque vlPFC results in selective object recognition deficits for specific images that most likely depend on recurrent computations. Their results implicate vlPFC, a recurrently connected circuit node, as critical to producing behaviorally sufficient object representations in the primate ventral visual stream.