Using decoders to understand working memory representations of 3D space in primate prefrontal neuronal ensembles

• DOI: 10.1167/jov.20.11.1474
• Publication Date: 2020-10-26T17:35:33Z
• Authors: MILAD KHAKI; MEGAN ROUSSY; NASIM MORTAZAVI; ROGELIO LUNA; ADAM SACHS; JULIO MARTINEZ-TR...
• Keywords: Neural decoding

Neurons in the primate lateral prefrontal cortex (LPFC) encode and maintain WM representations absence of external stimuli. Neural computations underlying spatial primates are traditionally studied using highly controlled tasks consisting simple 2D visual stimuli require a saccadic response. Therefore, there is little known about how populations LPFC neurons may transform 3D space for animals to navigate towards remembered object locations. To explore this issue, we created task that takes place virtual environment. In task, target presented one nine locations arena. The disappears during two-second delay period, then subject required location joystick. recordings were conducted two male rhesus macaques 10×10 Utah arrays located (area 8A), resulting total 3847 neurons. Using novel high-efficiency classification technique, decoded on single trial basis different epochs. This method resulted high decoding accuracy minimum number containing greatest amount target-specific information. Ensembles 8-12 60%-90% (chance= ~11). Importantly, also able decode outcomes (correct or incorrect). Furthermore, elucidate mechanism which neural ensembles space, each based depth direction (i.e. front, middle, back, right, center, left). results indicate retained with similar by ensembles. Moreover, show vary information they regarding either space.