Processing multimodal neural data and decoding neural dynamics using cross-modality inference

Our toolkits for investigating the complex functional and anatomical networks of the brain have been transformed as a result of recent advancements in neural recording and stimulation techniques, and gene sequencing of different types of brain cells. However, the methods that are currently in use have restrictions on either their spatial or temporal resolutions. As a consequence, there are gaps in our ability to monitor the activities of neuronal networks across the broad spatiotemporal scale of the brain, which spans from nanometers to millimeters and from microseconds to days. Development of transparent and flexible microelectrode arrays, which have made it possible to conduct multimodal studies, can link multiple spatial and temporal scales. Here, we present neural data analysis methods for a variety of multimodal studies that were carried out by combining techniques from the fields of electrophysiology and optical imaging. We demonstrate that distinct modalities convey a wealth of shared information that can be utilized to perform cross-modality inference and decode brain dynamics at various local and global scales. Such inferences could subserve the exploration of complex nonlinear mechanisms that exist between local field potentials and intracellular activities and ultimately pave the way toward less invasive brain-machine interfaces and more effective treatments for neurological disorders.