Ultra wideband for wireless real-time monitoring of neural signals

Abstract

Performance of an ultra wideband (UWB) wireless system for real-time neural signal monitoring is evaluated by comparing spiking characteristics between transmitted and received signals for different experimental set-ups. Spike detection quality is selected as the main spiking characteristic of evaluated signals. Results are presented in receiver-operating characteristics and area-under-the-curve (AUC). In order to assess spike detection quality, a set of artificially generated neural signals is constructed from real neural recordings such that the ground truth is known. Data analysis shows how channel signal-to-noise-ratio (SNR) variation affects AUC in different signal SNR cases. Signals with low SNRs get less affected by reduced channel SNRs than those with higher SNR. Increasing bit error rate modifies spiking characteristics such that an under-estimation of the spiking frequency occurs due to spike losses. For practical application of real-time neural signal monitoring, UWB seems to offer best transmission conditions in a near-body environment.

Real-time signal detection and classification algorithms for body-centered systems

“…A special characteristic of SVM is that the solution to a classification problem is represented by the support vectors that determine the maximum margin hyperplane. Given that the margin equals 2 w , maximizing the margins is equivalent as minimizing 2 2 w , thus the problem of finding the optimum H can be described by equation (2)(3)(4)(5).…”

“…The general link budget formula for an implant's and for the on-body device wireless links are equation (2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20) and (2-21) respectively:…”

“…Basic spike detection algorithms apply threshold-based detection to identify spikes and, although simple thresholding is attractive for real-time implementations because of its computational simplicity, it is thought to be sensitive to noise and requires user input to set effective threshold levels, as it is exposed in [4]. In neuronal recordings, Signal to Noise Ratios (SNR) vary with electrode geometry, size and position with respect to the target neuron [5].…”