An FSCV Deep Neural Network: Development, Pruning, and Acceleration on an FPGA

Fast-scan cyclic voltammetry (FSCV) is an electrochemical technique for measuring rapid changes in the extracellular concentration of neurotransmitters within the brain. Due to its fast scan rate and large output-data size, the current analysis of the FSCV data is often conducted on a computer external to the FSCV device. Moreover, the analysis is semi-automated and requires a good understanding of the characteristics of the underlying chemistry to interpret, making it unsuitable for real-time implementation on low-resource FSCV devices. This paper presents a hardware-software co-design approach for the analysis of FSCV data. Firstly, a deep neural network (DNN) is developed to predict the concentration of a dopamine solution and identify the data recording electrode. Secondly, the DNN is pruned to decrease its computation complexity, and a custom overlay is developed to implement the pruned DNN on a low-resource FPGA-based platform. The pruned DNN attains a recognition accuracy of 97.2% with a compression ratio of 3.18. When the DNN overlay is implemented on a PYNQ-Z2 platform, it achieves the execution time of 13 ms and power consumption of 1.479 W on the entire PYNQ-Z2 board. This study demonstrates the possibility of operating the DNN for FSCV data analysis on portable FPGA-based platforms.

I. Introduction

Fast-scan cyclic voltammetry (FSCV) is a high-speed electrochemical technique that has been largely used for the detection and measurement of in-vivo neurotransmitters, e.g., changing concentrations of dopamine, in real-time [1], [2]. By applying FSCV at carbon-fiber microelectrodes, it is possible to reveal direct information about neurochemical changes that may be relevant to neurological disease states. As shown in Fig. 1, the voltage present at a carbon-fiber microelectrode is rapidly changed to oxidize dopamine and reduce dopamine-O-quinone, often using a preconcentration step to increase sensitivity [3]. The resulting peak current is used to find the instantaneous concentration of dopamine. The key advantages of the FSCV technique include its high degree of sensitivity, as well as excellent temporal and spatial resolution [4], [5]. However, as FSCV is typically used with unmodified electrodes, the signals of analytes of interest can be electrode dependent. Fig. 1.

FSCV waveform for dopamine. Scanning electron micrograph (SEM) of carbon-fiber microelectrode (7 μm in diameter), with the dopamine redox reactions. Dopamine is oxidized to form dopamine-o-quinone, which can be reduced at the electrode back to dopamine [3].