A 4096 channel event-based multielectrode array with asynchronous outputs compatible with neuromorphic processors

Abstract Bio-signal sensing represents a pivotal domain in the medical applications of bioelectronics. Traditional methods have, so far, focused on capturing these signals as accurately as possible, leading to high sampling rates in clocked synchronous architectures. Given the sparse activity of bio-signals, this approach often results in large amounts of digitized data with no relevant information and in a significant amount of energy consumed during transmission. Here, we introduce a self-clocked microelectrode array (MEA) that senses and digitizes bio-signals at the pixel level by encoding their changes as asynchronous digital address-events, significantly reducing the amount of data that needs to be transmitted off-chip. This novel MEA comprises an array of 64 × 64 electrodes, an asynchronous 2D-arbiter, and an Address-Event Representation (AER) communication block. Each pixel within the array operates autonomously, monitoring input signals for relative voltage fluctuations instigated by cellular activity. Upon detecting a sufficiently large signal change, the pixel produces a digital pulse encoded with its corresponding address. Positive signal changes are encoded as “up” events, while negative ones are encoded as “down” events and, upon generation, are routed off-chip instantly via the asynchronous arbiter. Here, we present results from the chip characterization and experimental measurements using electrogenic cells. Moreover, we interface the MEA to a mixed-signal neuromorphic processor, demonstrating a prototype for end-to-end event-based bio-signal sensing and processing.