Mapping eye, arm, and reward information in frontal motor cortices using ECoG in nonhuman primates

• ECoG in nonhuman primates has been used to map eye, arm, and reward information in frontal motor cortices.¹
• The work directly informs invasive motor and cognitive BMI decoder design and neural decoding methods (Journal of Neuroscience).^{1 1}

Weekly enrichment (2026-07-20)

• The study used high-density micro-electrocorticography (µECoG) in two male rhesus macaques (Monkey 1 “Beignet,” 9 years, 11.3 kg; Monkey 2 “Affogato,” 10 years, 10.6 kg) to map eye, arm, and reward signals over a ~1.37 cm² area spanning primary motor cortex, premotor cortex, frontal eye field, and dorsolateral prefrontal cortex.^{2 3}
• Neural signals were recorded with a 244-electrode µECoG array (762 µm inter-electrode pitch, 229 µm contact size) embedded in a silicone artificial dura; 240 channels were amplified and digitized at 25 kHz using an eCube system (White-Matter Inc., Seattle).³
• Target direction (8 peripheral targets) was decoded with linear discriminant analysis over six spectral bands (delta 0.1–4 Hz, theta 4–8 Hz, alpha 8–14 Hz, beta 14–30 Hz, gamma 30–80 Hz, high-gamma 80–150 Hz) using 5-fold cross-validation, with chance level at 12.5% (100/8).^{2 3}
• Mean single-channel decoding accuracy in Monkey 1 rose across task phases: 13.8 ± 1.46% at target onset, 17.7 ± 3.26% at saccade onset, 17.5 ± 3.23% at reach onset, and 19.2 ± 4.46% at reward.²
• After behavioral trial selection, 533 of 800 successful trials were retained for Monkey 1 and 829 of 1,078 for Monkey 2; channel quality screening kept 238/244 channels (Monkey 1) and 233/244 (Monkey 2).²
• Arm-movement decoding from motor cortex was measurably degraded by task-irrelevant eye movements, providing direct evidence that eye and arm information are spatially intermixed across motor cortex.^{2 3}
• Intertrial phase clustering localized reward-related activity primarily around the principal sulcus (dorsolateral prefrontal cortex) and near the arcuate sulcus (premotor cortex).^{2 3}
• The paper appeared in the Journal of Neuroscience (vol. 45, issue 12, e1536242025; published Jan 29, 2025; University of Washington) and frames µECoG as a route to multi-signal, eye/reward-aware brain–computer interfaces (DOI 10.1523/jneurosci.1536-24.2025).²
• Context: earlier high-density ECoG over macaque ventral premotor cortex could predict reach-return onset and direction, with high-gamma most predictive during movement and mid-band beta/low-gamma more predictive before movement onset—consistent with this study’s band-specific decoding.⁴

Footnotes

1. https://news.google.com/rss/articles/CBMiYEFVX3lxTFBpVXRDMTZFMVF6T21kSmx0SXlwS1pTa3NFR3VPMXNqcDJMTUNBR0VXN00tVmV4VG1Ndkc0eWVUR1YtVzVyd1JwSWZmR3YtbXpxSEQ2cUdaOS0xZHZyUWJFbA?oc=5 ↩ ↩² ↩³

2. https://www.jneurosci.org/content/45/12/e1536242025 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

3. https://pmc.ncbi.nlm.nih.gov/articles/PMC11343120/ ↩ ↩² ↩³ ↩⁴ ↩⁵

4. https://doi.org/10.1038/s41598-018-28940-7 ↩