Neural decoding articles from across Nature Portfolio

A brain–computer interface can decode semantic representations of perceived and imagined stimuli at the level of words and phrases from functional MRI recordings of brain activity.

Using long-term brain recordings in patients with chronic pain, we identified objective biomarkers of real-world subjective pain intensity over many months. Spontaneous chronic pain states were predicted most reliably by sustained changes in the activity of the orbitofrontal cortex, whereas acute pain was most associated with signals from the anterior cingulate cortex.

Bidirectional neuromorphic brain interfaces, based on interconnecting brain networks with artificial spiking neural networks, aim to optimize neural prostheses to help people with paralysis to regain autonomy. Their emergence questions the concept of human subjectivation, the continuous process by which we become and remain the subject of our life.

The representational geometry of neural population activity in the somatosensory cortex of mice allows for high flexibility needed to perform complex tasks and for generalization to novel tasks at the same time.

To interact safely with our environment, we must be able to judge our confidence in what we perceive. But what cues do we use to compute perceptual confidence? Geurts et al.¹ decode brain activity and show that perceptual confidence is based on the distribution of sensory uncertainty, combining uncertainty driven by the input and the visual system.

Sensory information encoding in the mouse brain is more suboptimal when mice make correct decisions than when they make incorrect ones. These suboptimal encoding structures can help information flow between different brain regions, enhancing the ability of these brain regions to work together to make decisions.