Abstract
Characterizing, monitoring, and controlling neural dynamics remains an elusive goal for research efforts ranging from neuroscience to computer science. The approaches attempt to unearth a richer interpretation for the nonlinear, time-varying hierarchy of biological hardware organizations. Some promising advances with specific investigations focusing on classes of problems pertaining to how emergent neural network dynamics produces 1) distinct cognitive states including degenerative pathologies and 2) reproducing, maintaining and controlling efficient information processing capabilities in algorithms or addressing needed innovations in power-hungry computing hardware. Higher order neural global dynamical features (e.g., pathological, or healthy states) are a manifestation of the cumulative interactions of its local components (e.g., populations of neural and/or non-neuronal cells and their interactions). Hence, a thorough investigation of distinct biological mediums of coupling such as synaptic interactions between neural cells has provided insight on the higher-order formation, maintenance, or pruning of memory (a global network feature). These dynamical transformations (or information-processing mediums) are anchored by the local reinforcement or weakening of synaptic strength/weight (local network interactions). Such characterization enables identifying different targets of intervention to modulate neural states (control). Furthermore, this helps to better illustrate how local configurations of neural building blocks can be coupled to produce higher-order dynamical signatures. These can include collaborative or competitive transient transformations underpinning the core of information processing in complex networks (neural, artificial, or otherwise) responsible for the creation, preservation, or elimination of information. In other words, a current goal is to find the local neural components which carry the most weight in steering global emergent network features towards desirable operating states (from administering therapeutic interventions to controlling information processing such as memory encoding/recall). Simply stated, the objective is to find the most efficient targets for control. Hence, it’s needed to experimentally reproduce higher levels of neural organization which can undergo rigorous experimental testing to further explore efficient local targets for control. Therefore, Three-dimensional (3D) organoid engineering aims to steer cell aggregates toward physiological mimicking of human tissue and organ systems at the cellular level, essentially serving as tissue and organ proxies that recapitulate biological parameters (e.g., spatial organization of heterogeneous tissue-specific cells, cell-cell interactions, etc.). Currently, attempts at generation of brain organoids do not mature beyond the prenatal brain equivalent, the major obstacle being the lack of vascularization in the initial embryoid bodies that ultimately limit the growth and maturation of the organoids. For example, in vitro culture conditions limit the size of the organoid, neuronal maturation, and subsequent production of more complete cell types, such as astrocytes and oligodendrocytes (misrepresenting the full scale, heterogenous complexity of a brain network). Thus, attention is turned toward the generation of a brain-on-a-chip model that can serve as a relevant model of the human brain in its recapitulation of the neuronal circuit (i.e., organoid-on-chip or “OOC”; brain-on-chip or “BOC”). This platform houses 3D neural organoids enabling detection of electrical activity dynamics under various external conditions. Detection of electrical activity on a 3D spatial organoid provides a precise global quantification of the emergent, global neural dynamics at this particular scale. Furthermore, modification of this activity can be detected with respective changes to the external environment. These novel approaches address the current limitations regarding organoid engineering, maturation, and experimental testing for producing a richer complexity of neural circuit organization in vitro enabling greater degrees of interventional experimentation for detailed observation, manipulation, and control of the global network states. Detected, global voltage excitations of organoids can be used as a global constraint from which local components of the network and their time-varying interactions can be inferred upon through theoretical statistical mechanics. That is, the underlying interactions producing this emergent, mesoscopic electrical activity (neural organoid voltage excitations) provides a global constraint from which the underlying interactions of neural and non-neuronal cells (astrocytes) can be theoretically derived providing additional insight towards the degree of influence each constituent (network node) type has on the global network state.
