Search

Neurons are the basic blocks in the Central Nervous System (CNS). Simulation and hardware realization of these blocks are vital in neuromorphic engineering. This paper presents a set of multiplierless mathematical equations based on $2^{X}$ terms to achieve a low-cost, high-speed, and high-accuracy digital implementation of Hodgkin-Huxley (HH) neuron model. The HH model is the most complicated and high-accuracy among the mathematical neuron models. The proposed model can reproduce spiking behaviors of the original HH model with high precision. To validate the mathematical simulation results, the proposed model has been synthesized and implemented on Field-Programmable Gate Array (FPGA) development board. Hardware synthesis and physical implementations reveal that the biological behavior of different spiking patterns can be reproduced with higher performance and significantly lower implementation costs compared with the original HH model. Also, in this approach the maximum frequency of 200 MHz is achievable which is valuable in comparison with other similar works. [ABSTRACT FROM AUTHOR]

The construction of a design matrix is critical to the accurate detection of activation regions of the brain in functional magnetic resonance imaging (fMRI). The design matrix should be flexible to capture the unknown slowly varying drifts as well as robust enough to avoid overfitting. In this paper, a sparse Bayesian learning method is proposed to determine a suitable design matrix for fMRI data analysis. Based on a generalized linear model, this learning method lets the data itself determine the form of the regressors in the design matrix. It automatically finds those regressors that are relevant to the generation of the fMRI data and discards the others that are irrelevant. The proposed approach integrates the advantages of currently employed methods of fMRI data analysis (the model-driven and the data-driven methods). Results from the simulation studies clearly reveal the superiority of the proposed scheme to the conventional t-test method of fMRI data analysis. [ABSTRACT FROM AUTHOR]

In this paper, we present a fully integrated front-end of a portable spectroreflectometry-based brain imaging system dedicated for acquisition of modulated optical signals at a frequency of 1 Hz to 25 kHz. The proposed front-end preamplifier is composed of a photodetector, a transimpedance preamplifier, a two-stage voltage amplifier and a mixer. Strict constraints regarding noise thus have to be considered. The preamplifier consists of a transimpedance block featuring a 95-dBΩ gain and an average input current noise density at the frequency of interest of approximately 3 pA/√Hz. Each of the two subsequent voltage amplifiers allows the user to obtain an additional 25-dB gain. Considering the tuning capabilities and the losses due to the filters and the nonideal buffers, the proposed front-end allows us to obtain a total gain up to 145 dB. The back-end of the amplification chain is composed of a mixer which is used to produce a continuous voltage proportional to the amplitude of the input optical signals. All those features were integrated using CMOS 0.18-μm technology and the experimental results are in agreement with the initial design requirements. [ABSTRACT FROM AUTHOR]