Back

Investigating the mechanisms underlying saccade generation in the frontal eye fields using multi-site microstimulation

Johnston, R.; Konecky, R.; Katnani, H. A.; Gandhi, N. J.; Smith, M. A.

2025-12-29 neuroscience
10.64898/2025.12.29.696832 bioRxiv
Show abstract

The frontal eye field (FEF), a region of frontal cortex, has long been associated with the cortical control of eye movements. Classically, saccades can be reliably evoked by delivering low-intensity electrical microstimulation to the FEF. Although this makes clear the importance of FEF in the descending control of eye movements, the way in which population activity in the FEF is integrated by downstream regions to generate a motor command remains a mystery. To probe these mechanisms, we used a 16-channel microelectrode array to deliver microstimulation to the FEF of two awake, behaving monkeys. First, we found that larger current intensities were required to evoke changes in saccade direction relative to saccade amplitude when single-site saccades were evoked by stimulating a single contact on the array. Second, when stimulating two contacts simultaneously to investigate how population activity in the FEF is read out, a new polar average model more accurately predicted the amplitude and direction of dual-site saccades than traditional vector sum and vector average models. Using preexisting data from the superior colliculus (SC), we found that although the polar average model was more accurate at predicting saccade amplitude in the SC, it was no more accurate than traditional models at predicting saccade direction. Finally, when stimulating two contacts in FEF simultaneously with unequal current intensities, model accuracy depended on the amplitude of the saccades evoked by stimulating each individual site alone, suggesting that the brain may flexibly combine amplitude and direction information from the FEF to generate saccadic plans.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.