Theory & Data Analysis Tools

Sleep is critical to memory and learning. During rapid eye movement (REM) or non-REM (NREM) sleep, subgroups of cell assemblies in hippocampal and sensory cortical circuits are reactivated in a temporally coordinated manner, forming a cortical-hippocampal-cortical loop of information processing during memory consolidation. Deciphering neural codes of hippocampal-neocortical memories during sleep would reveal important circuit mechanisms of memory consolidation.

There have been remarkable advances in imaging technology, used routinely and pervasively in many human studies, that non-invasively measures human brain structure and function. Diffusion magnetic resonance imaging (dMRI) and structural MRI (sMRI) are used to infer locations of millions of interconnected white matter fiber tracts-known as the brain connectome-that act as highways for neural activity and communication across the brain.

Learning from feedback in the real w'orld is limited by constant fluctuations in reward outcomes associated with choosing certain options or actions. Some of these fluctuations are caused by fundamental changes in the reward values of those options/actions that necessitate dramatic adjustments to the current learning strategies, like in epiphany learning or one-shot learning [Chen & Krajbich, 2017; Lee et al. 2015]. Other changes represent inherent stochasticity in an otherwise stable environment and should be tolerated and ignored to maintain stable choice preferences.

Dendritic spines mediate essentially all excitatory connections and are thus critical elements in the brain but their function is still poorly understood. In particular, a key question is whether or not they are electrical compartments. To explore this, researchers have used cable theory and Goldman-Hodgkin-Huxley-Katz models, which form a theoretical foundation responsible for many cornerstone advances in neuroscience. However, these theories break down when applied to small neuronal compartments, such as dendritic spines, because they assume spatial and ionic homogeneity.

Real-time neural decoding centers on predicting behavior variables based on neural activity data, where the prediction is performed at a pace that reliably keeps up with the speed of the activity that is being monitored. Neuromodulation devices are becoming one of the most powerful tools for the treatment of brain disorders, enhancing neurocognitive performance, and demonstrating causality (Bergmann et al., 2016; Knotkova and Rasche, 2015). A precise neuromodulation system (Figure 1) integrates neural activity monitoring, real-time neural decoding, and neuromodulation.

Collaborating Pis and Consultant United States Pl: Anitha Pasupathy, Dept. of Biological Structure, University of Washington, Seattle, USA Co-Pl: Wyeth Bair, Dept. of Biological Structure, University of Washington, Seattle, USA Japan Pl: lsamu Motoyoshi, Dept.

Dopamine (DA) neurons are fundamental to many aspects of behavior, and dysfunction of the DA system contributes to a wide range of disorders, including drug addiction. How does DA contribute to such a diversity of functions and dysfunctions? Part of the answer may relate to recent discoveries that DA neurons respond to a wide range of behavioral variables - not only to reward and reward-predicting cues, as traditionally examined, but also to other variables including position, movement, and behavioral choices.

There is a fundamental gap in our understanding of the computational principles and neural mechanisms by which neural circuits represent complex objects like faces. This conceptual gap constitutes an important problem because, until it is filled, we will not be able to understand face recognition and the reasons for face blindness. The long-term goal is to understand the computational principles and neural mechanisms of face recognition and create a computer face-recognition system based on these principles.

Humans and other animals can choose their actions using multiple learning algorithms and decision­ making strategies. For example, habitual behaviors adapted to a stable environment can be selected using so-called model-free reinforcement learning algorithms, in which the value of each action is incrementally updated according to the amount of unexpected reward. The underlying neural mechanisms for this type of reinforcement learning have been intensively studied.