EOSFET

Abstract: Advances in neuronal probe technology to record brain activity have posed a significant challenge in performing necessary processing and analysis of the recorded data. To be able to infer meaningful conclusions from the recorded signals through these probes, sophisticated signal processing and analysis tools are required. This paper presents a MATLAB-based novel tool, ‘SigMate’, capable of performing various processing and analysis incorporating the available standard tools and our in-house custom tools. The present features include, data display (2D and 3D), baseline correction, stimulus artifact removal, noise characterization, file operations (file splitter, file concatenator, and file column rearranger), latency estimation, determination of cortical layer activation order, spike detection, spike sorting, and are gradually growing. This tool has been tested extensively for the recordings using the standard micropipettes as well as implantable neural probes based on EOSFETs (Electrolyte-Oxide-Semiconductor Field Effect Transistors) and will be made available to the community shortly.

Abstract: In recent years the experimental toolkit at disposal of neuroscientists to investigate electrophysiologically the brain “in vivo” down to the level of neuronal microcircuits circuits and to elucidate their fundamental mechanisms for mapping and processing information has grown rapidly and even beyond expectations [1]. Driven by the compelling need of recording large numbers of neurons within the cortex and deeper structures, in a minimally invasive manner and over long time periods [2–4], the development of implantable brain probes based on microelectromechanical systems (MEMS) with arrays of microelectrodes has experienced a significant boost, leading to substantial optimization of pioneering approaches conceived in the 1970s [5] as well as to the development of novel technologies. Multielectrode arrays (MEAs) and multitransistor arrays (MTAs) integrated in silicon microchips constitute two major representatives from this class of brain implantable probes. Originally developed as “in vitro” prototypes for recording dissociated neurons or brain slices and other excitable cells [5–7], MEA and MTA reflect two different philosophies for transducing a neuronal electrical signal to a semiconductor chip, that is, either through a metal microelectrode or by means of an electrolyte–oxide–semiconductor field-effect transistor (EOSFET), a modified version of the metal–oxide–semiconductor field-effect transistor (MOSFET) that is widely used in integrated circuits [8] (Fig. 8.1).

Abstract: Existing brain-machine interfacing techniques allow either high precision recordings from one or a few single neurons, or low spatial resolution recordings with a sparse sampling within the networks. Through our app-roach an efficient simultaneous bidirectional communication to the brain is realized using capacitively coupled recording and stimulation sites arranged in a large 2D multi-transistor array (MTA) with 1000 elements, integrated to a planar chip at high resolution (10μm pitch and below). The aim of the present work is to evaluate the reliability of a simple-generation silicon micro-device in recording neuronal signals from rat brain. Simultaneous recording of signals using this chip from the somatosensory cortex (S1) of living rat, are compared to standard in vivo recordings with a glass micropipette. We show that the two types of signals are identical, indicating the possibility to record signals at the same time from different sites and to perform a real-time electrical imaging of the brain cortex in vivo.

Abstract: Brain-machine interfaces are currently based on techniques allowing either to record at high resolution from one or a few single neurons, or low spatial resolution recordings with a sparse sampling within the networks. To better interface to circuitries and to understand their role in sensory systems or cognition, higher resolution probes are required. In this paper we report a novel technique capable of recording cortical signals at a high resolution providing an electrical imaging of the cortical region under examination. Imaging was performed using two different types of electrolyte-(metal)-oxide-semiconductor field effect transistor, E(M)OSFET based multi-transistor arrays (MTAs): 1) 64 recording elements, integrated into a planar chip at high resolution (pitch: 30 μm-40 μm); 2) a matrix of 128 × 128 recording elements, integrated at a higher resolution (pitch: 7.4 μm, type: EMOSFET). These silicon micro-devices were capable of simultaneous recording of neuronal signals from the somatosensory cortex (S1) of the rat brain and were suitable in performing a real-time electrical imaging of the brain cortex in-vivo.