Towards an Accurate Identification of Pyloric Neuron Activity with VSDi

Abstract

Abstract. Voltage-sensitive dye imaging (VSDi) which enables simultaneous optical recording of many neurons in the pyloric circuit of the stomatogastric ganglion is an important technique to supplement electrophysiological recordings. However, utilising the technique to identify pyloric neurons directly is a computationally exacting task that requires the development of sophisticated signal processing procedures to analyse the tri-phasic pyloric patterns generated by these neurons. This paper presents our work towards commissioning such procedures. The results achieved to date are most encouraging. .

A robust data-driven approach to the decoding of pyloric neuron activity

“…First, the multiresolution procedure described in [12] was applied to the individually demarcated ROIs in order for the relevant signal sub/components of the PR, if present, to be accurately extracted. Further, the signal components bearing the first and second harmomics of the pyloric frequency must also be isolated, to enable estimation of the duty cycle and thus identification of the participating pyloric neurons [13]. Second, for the extracellularly captured data on the lvn, the primary task is to separate the individual phase of the PR from each computed pyloric cycle, in accordance with the spiking/activity pattern characteristic of the respective participating neuron.…”

“…The deconstruction of the highly structured and regular phasing of the pyloric activities recorded on the lvn is constructed using the widely studied multi-resolution technique of time-frequency analysis [13,15]. Importantly, this was assisted by the s-SSA procedure described above, which sought to remove noise from the lvn whilst selectively preserving the pyloric frequency and its higher order harmonics.…”

“…This was achieved by selecting a frequency band in the constructed spectrogram where the amplitude response produced in the PY-timed pyloric cycle and the quiescent phase/period are significantly lower (~75%) than that in either of the LP-timed and PD-timed pyloric cycle. Using a simple thresholding procedure, the respective time periods of PY phase were segmented from the individual pyloric cycle recorded on the lvn and, subsequently, the corresponding duty cycle of the PY phase was computed [13].…”

“…The latter were developed to compare and correlate the optically recorded intensity of the fluorescence produced by the voltage-sensitive dye (Di-4-ANEPPS). The dye was bath-applied exogenously to the STG sample and the firing phases deconstructed from the activity patterns captured extracellularly through the lateral ventricular nerve (lvn) [12]- [13]. Specifically, the biologically meaningful and computationally well-defined duty cycle of the participating neurons was used as a biomarker to help identify the PY neuron/cell from the temporally captured sequence of images of the bath dyed STG sample [13].…”

“…The dye was bath-applied exogenously to the STG sample and the firing phases deconstructed from the activity patterns captured extracellularly through the lateral ventricular nerve (lvn) [12]- [13]. Specifically, the biologically meaningful and computationally well-defined duty cycle of the participating neurons was used as a biomarker to help identify the PY neuron/cell from the temporally captured sequence of images of the bath dyed STG sample [13]. This paper extends the above analysis of the pyloric constrictor neuron (PY, x5) to consider lateral pyloric (LP, x1) and pyloric dilator (PD, x2) neurons based on the rhythmic and triphasic activity patterns routinely observable from the lvn recordings.…”