Functional (ir)relevance of posterior parietal cortex during audiovisual change detection.

The posterior parietal cortex (PPC) plays a key role in integrating sensory inputs from different modalities to support adaptive behavior. Neuronal activity in PPC reflects perceptual decision making across behavioral tasks, but the mechanistic involvement of PPC is unclear. In an audiovisual change detection task, we tested the hypothesis that PPC is required to arbitrate between the noisy inputs from the two different modalities and help decide in which modality a sensory change occurred. In trained male mice, we found extensive single-neuron and population-level encoding of task-relevant visual and auditory stimuli, trial history, as well as upcoming behavioral responses. However, despite these rich neural correlates, which would theoretically be sufficient to solve the task, optogenetic inactivation of PPC did not affect visual or auditory performance. Thus, in spite of neural correlates faithfully tracking sensory variables and predicting behavioral responses, PPC was not relevant for audiovisual change detection. This functional dissociation questions the role of sensory- and task-related activity in parietal associative circuits during audiovisual change detection. Furthermore, our results highlight the necessity to dissociate functional correlates from mechanistic involvement when exploring the neural basis of perception and behavior.SIGNIFICANCE STATEMENTThe Posterior Parietal Cortex (PPC) is active during many daily tasks, but capturing its function has remained challenging. Specifically, it is proposed to function as an integration hub for multisensory inputs. Here, we tested the hypothesis that, rather than classical cue integration, mouse PPC is involved in the segregation and discrimination of sensory modalities. Surprisingly, even though neural activity tracked current and past sensory stimuli and reflected the ongoing decision-making process, optogenetic inactivation did not affect task performance. Thus, we show an apparent redundancy of sensory and task-related activity in mouse PPC. These results narrow down the function of parietal circuits, as well as direct the search for those neural dynamics that causally drive perceptual decision making.

Non-REM sleep and the neural correlates of consciousness: more than meets the eyes.

The scientific study of the neural correlates of consciousness (NCC) has long relied on comparing conditions in which consciousness is normally present with others in which it is impaired. Brain lesions offer a unique opportunity to understand which anatomical networks are needed to sustain consciousness, but provide limited insights on the patterns of neural activity that can support conscious processing. Non-REM sleep, on the other hand, has long epitomized the typical case of a non-conscious yet fully active brain. Consequently, the differences in neural activity existing between wakefulness and non-REM sleep have also been used to define the NCC. Recently, however, several studies have started challenging the traditional understanding of neuronal activity during wakefulness and sleep. First, oscillatory dynamics characteristic of non-REM sleep - such as slow oscillations - have been reported to occur during wakefulness. Second, neural dynamics typical of conscious states have also been observed during non-REM sleep. Finally, the disconnection in cortical effective connectivity that has been indicated as one of the hallmarks of the loss of consciousness that occurs during non-REM sleep has recently been shown to be a less general phenomenon than previously thought. Here I will provide an overview of these recent findings, and discuss their implications for understanding the real nature of the NCC.

A simple highly efficient non invasive EMG-based HMI.

Muscle activity recorded non-invasively is sufficient to control a mobile robot if it is used in combination with an algorithm for its asynchronous analysis. In this paper, we show that several subjects successfully can control the movements of a robot in a structured environment made up of six rooms by contracting two different muscles using a simple algorithm. After a small training period, subjects were able to control the robot with performances comparable to those achieved manually controlling the robot.