Distinct context- and content-dependent population codes in superior colliculus during sensation and action

Abstract

Sensorimotor transformation is the process of first sensing an object in the environment and then producing a movement in response to that stimulus. For visually guided saccades, neurons in the superior colliculus (SC) emit a burst of spikes to register the appearance of stimulus, and many of the same neurons discharge another burst to initiate the eye movement. We investigated whether the neural signatures of sensation and action in SC depend on context. Spiking activity along the dorsoventral axis was recorded with a laminar probe as Rhesus monkeys generated saccades to the same stimulus location in tasks that require either executive control to delay saccade onset until permission is granted or the production of an immediate response to a target whose onset is predictable. Using dimensionality reduction and discriminability methods, we show that the subspaces occupied during the visual and motor epochs were both distinct within each task and differentiable across tasks. Single-unit analyses, in contrast, show that the movement-related activity of SC neurons was not different between tasks. These results demonstrate that statistical features in neural activity of simultaneously recorded ensembles provide more insight than single neurons. They also indicate that cognitive processes associated with task requirements are multiplexed in SC population activity during both sensation and action and that downstream structures could use this activity to extract context. Additionally, the entire manifolds associated with sensory and motor responses, respectively, may be larger than the subspaces explored within a certain set of experiments.

Multiunit frontal eye field activity codes the visuomotor transformation, but not gaze prediction or target memory, in a delayed saccade task

“…Recently, there has been increased interest in comparing SU and MU activity in both animals and humans (Ayar et al, 2023;Ahmadi et al, 2021;Burns et al, 2010;Christie et al, 2015;Drebitz et al, 2019;Gilja et al, 2015;Pandarinath et al, 2017;Trautmann et al, 2019). For example, Trautmann and colleagues (Trautmann et al, 2019) suggested that MU recordings provide both the practical benefit eliminating spike sorting (required for SU analysis) and reducing the dimensionality of the signal without losing important information content.…”

“…Further, whereas the spatiotemporal progression of the SU data showed various dips and rises during the delay period, perhaps related to microsaccades (Heusser et al, 2022), the MU data showed a 'smoother' transition, suggesting a more robust sensorimotor signature. Considering the relative ease of recording online MU activity (without the need for offline spike sorting), this suggests that MU activity could be sufficient for certain practical applications requiring low-dimensional information (Ayar et al, 2023;Trautmann et al, 2019). For example, MU activity could be observed to rapidly assess the health of sensorimotor transformations during presurgical recordings in patient populations.…”

“…Traditionally, neuroscientists have isolated single-unit (SU) action potentials of individual neurons from raw recorded background noise, based upon its unique spike shape; a process called spike-sorting. Recently, however, there has been increased interest in the use of multiunit (MU) activity (i.e., unsorted signals recorded from the tip of an electrode, which may arise from several nearby neurons (Ayar et al, 2023;Buzsáki, 2004;Panzeri et al, 2015; Quian Quiroga and Panzeri, 2009;Rossi-Pool and Romo, 2019). It has been suggested that MU activity recordings have practical and theoretical implications for experimental and applied electrophysiology (Ahmadi et al, 2021;Choi et al, 2010;Fraser et al, 2009;Leuthardt et al, 2021;Mattia et al, 2010;Sharma et al, 2015;Smith et al, 2013;Trautmann et al, 2019;Zhang and Constandinou, 2021).…”

“…The process of extracting / sorting MU activity is more reproducible since it does not eliminate data or depend on specific sorting techniques / parameters. Some studies found that MU activity provided a more stable signal with higher spatial resolution than SU activity (Ayar et al, 2023;Drebitz et al, 2019;Lewicki, 1998;Sharma et al, 2015;Xing et al, 2009). Further, MU activity profiles are similar to local field potential (LFP) signals but tend to show less correlation between nearby neurons (Ahmadi et al, 2021;Burns et al, 2010;Drebitz et al, 2019;Telenczuk and Destexhe, 2020).…”

“…Further, MU activity profiles are similar to local field potential (LFP) signals but tend to show less correlation between nearby neurons (Ahmadi et al, 2021;Burns et al, 2010;Drebitz et al, 2019;Telenczuk and Destexhe, 2020). Finally, MU analysis is capable of decoding the task and behavior by reducing the dimensionality of neural states (Ayar et al, 2023;Quian Quiroga and Panzeri, 2009;Rossi-Pool and Romo, 2019;Smith et al, 2013;Trautmann et al, 2019). Therefore, MU activity analysis has become popular for applications such as brain-machine interfaces, recording paralytic motor activity (Gilja et al, 2015;Pandarinath et al, 2017) and other clinical / preclinical trials (Christie et al, 2015;Fraser et al, 2009;Kao et al, 2017;Todorova et al, 2014).…”