Monkeys learn to report their own sensory cortical population activity

A groundbreaking study reveals that monkeys can learn to report their own activity in sensory cortical populations, a discovery that could shed light on the fundamental workings of sensory perception. By using a closed-loop, neurofeedback paradigm, researchers were able to train the monkeys to focus on specific patterns of activity in their own brain cells, recorded in real-time.

The experiment involved shaping stimuli and implanted electrode arrays in the visual area V4 of macaque monkeys. The monkeys were then trained to rely on their V4 activity along specific axes in neural state space, essentially learning to "think" about their own brain activity. Remarkably, the monkeys successfully increased the correlation between neuron choices and the trained axes in their neural population state space.

The findings suggest that closed-loop neural feedback can align neuron-choice beyond the limits of natural perceptual training. This misalignment, which has been a subject of controversy, may simply reflect constraints on learning within the naturally available training regimes. The experiment also confirmed that altering sensory readouts without disrupting stimulus selectivity or noise correlations had a positive impact on neuron-choice correlation.

In essence, this study demonstrates that monkeys can enhance their sensory perception by learning to control their own brain activity, opening up new avenues for understanding the complex interplay between perception and neural activity.