Neurochemical Concentration Prediction Using Deep Learning vs Principal Component Regression in Fast Scan Cyclic Voltammetry: A Comparison Study

Neurotransmitters, such as dopamine and serotonin, are responsible for mediating a wide array of neurologic functions, from memory to motivation. From measurements using fast scan cyclic voltammetry (FSCV), one of the main tools used to detect synaptic efflux of neurochemicals in vivo, principal component regression (PCR), has been commonly used to predict the identity and concentrations of neurotransmitters. However, the sensitivity and discrimination performance of PCR have room for improvement, especially for analyzing mixtures of similar oxidizable neurochemicals. Deep learning may be able to address these challenges. To date, there have been a few studies to apply machine learning to FSCV, but no attempt to apply deep learning to neurotransmitter mixture discrimination and no comparative study have been performed between PCR and deep learning methods to demonstrate which is more accurate for FSCV analysis so far. In this study, we compared the neurochemical identification and concentration estimation performance of PCR and deep learning in an analysis of FSCV recordings of catecholamine and indolamine neurotransmitters. Both analysis methods were tested on in vitro FSCV data with a single or mixture of neurotransmitters at the desired concentration. In addition, the estimation performance of PCR and deep learning was compared in incorporation with in vivo experiments to evaluate the practical usage. Pharmacological tests were also conducted to see whether deep learning would track the increased amount of catecholamine levels in the brain. Using conventional FSCV, we used five electrodes and recorded in vitro background-subtracted cyclic voltammograms from four neurotransmitters, dopamine, epinephrine, norepinephrine, and serotonin, with five concentrations of each substance, as well as various mixtures of the four analytes. The results showed that the identification accuracy errors were reduced 5-20% by using deep learning compared to using PCR for mixture analysis, and the two methods were comparable for single analyte analysis. The applied deep-learning-based method demonstrated not only higher identification accuracy but also better discrimination performance than PCR for mixtures of neurochemicals and even for in vivo testing. Therefore, we suggest that deep learning should be chosen as a more reliable tool to analyze FSCV data compared to conventional PCR methods although further work is still needed on developing complete validation procedures prior to widespread use.

Enhanced Dopamine Sensitivity Using Steered Fast-Scan Cyclic Voltammetry

Fast-scan cyclic voltammetry (FSCV) is a technique for measuring phasic release of neurotransmitters with millisecond temporal resolution. The current data are captured by carbon fiber microelectrodes, and non-Faradaic current is subtracted from the background current to extract the Faradaic redox current through a background subtraction algorithm. FSCV is able to measure neurotransmitter concentrations in vivo down to the nanomolar scale, making it a very robust and useful technique for probing neurotransmitter release dynamics and communication across neural networks. In this study, we describe a technique that can further lower the limit of detection of FSCV. By taking advantage of a "waveform steering" technique and by amplifying only the oxidation peak of dopamine to reduce noise fluctuations, we demonstrate the ability to measure dopamine concentrations down to 0.17 nM. Waveform steering is a technique to dynamically alter the input waveform to ensure that the background current remains stable over time. Specifically, the region of the input waveform in the vicinity of the dopamine oxidation potential (∼0.6 V) is kept flat. Thus, amplification of the input waveform will amplify only the Faradaic current, lowering the existing limit of detection for dopamine from 5.48 to 0.17 nM, a 32-fold reduction, and for serotonin, it lowers the limit of detection from 57.3 to 1.46 nM, a 39-fold reduction compared to conventional FSCV. Finally, the applicability of steered FSCV to in vivo dopamine detection was also demonstrated in this study. In conclusion, steered FSCV might be used as a neurochemical monitoring tool for enhancing detection sensitivity.

Feasibility of Applying Fourier Transform Electrochemical Impedance Spectroscopy in Fast Cyclic Square Wave Voltammetry for the In Vivo Measurement of Neurotransmitters

We previously reported on the use of fast cyclic square wave voltammetry (FCSWV) as a new voltammetric technique. Fourier transform electrochemical impedance spectroscopy (FTEIS) has recently been utilized to provide information that enables a detailed analytical description of an electrified interface. In this study, we report on attempts to combine FTEIS with FCSWV (FTEIS-FCSWV) and demonstrate the feasibility of FTEIS-FCSWV in the in vivo detection of neurotransmitters, thus giving a new type of electrochemical impedance information such as biofouling on the electrode surface. From FTEIS-FCSWV, three new equivalent circuit element voltammograms, consisting of charge-transfer resistance (R_ct), solution-resistance (R_s), and double-layer capacitance (C_dl) voltammograms were constructed and investigated in the phasic changes in dopamine (DA) concentrations. As a result, all R_ct, R_s, and C_dl voltammograms showed different DA redox patterns and linear trends for the DA concentration (R² > 0.99). Furthermore, the R_ct voltammogram in FTEIS-FCSWV showed lower limit of detection (21.6 ± 15.8 nM) than FSCV (35.8 ± 17.4 nM). FTEIS-FCSWV also showed significantly lower prediction errors than FSCV in selectivity evaluations of unknown mixtures of catecholamines. Finally, C_dl from FTEIS-FCSWV showed a significant relationship with fouling effect on the electrode surface by showing decreased DA sensitivity in both flow injection analysis experiment (r = 0.986) and in vivo experiments. Overall, this study demonstrates the feasibility of FTEIS-FCSWV, which could offer a new type of neurochemical spectroscopic information concerning electrochemical monitoring of neurotransmitters in the brain, and the ability to estimate the degree of sensitivity loss caused by biofouling on the electrode surface.

Sensitive and Selective Measurement of Serotonin in Vivo Using Fast Cyclic Square-Wave Voltammetry

Although N-shaped fast scan cyclic voltammetry (N-FSCV) is well-established as an electroanalytical method to measure extracellular serotonin concentrations in vivo, it is in need of improvement in both sensitivity and selectivity. Based on our previous studies using fast cyclic square-wave voltammetry (FCSWV) for in vivo dopamine measurements, we have modified this technique to optimize the detection of serotonin in vivo. A series of large amplitude square-shaped potentials was superimposed onto an N-shaped waveform to provide cycling through multiple redox reactions within the N-shaped waveform to enhance the sensitivity and selectivity to serotonin measurement when combined with a two-dimensional voltammogram. N-Shaped fast cyclic square-wave voltammetry (N-FCSWV) showed significantly higher sensitivity to serotonin compared to conventional N-FSCV. In addition, N-FCSWV showed better performance than conventional N-shaped FSCV in differentiating serotonin from its major interferents, dopamine and 5-hydroxyindoleascetic acid (5-HIAA). It was also confirmed that the large amplitude of the square waveform did not influence local neuronal activity, and it could monitor electrical stimulation evoked phasic release of serotonin in the rat substantia nigra pars reticulata (SNr) before and after systemic injection of escitalopram (ESCIT, 10 mg/kg i.p.), a serotonin selective reuptake inhibitor.

Tracking tonic dopamine levels in vivo using multiple cyclic square wave voltammetry

For over two decades, fast-scan cyclic voltammetry (FSCV) has served as a reliable analytical method for monitoring dopamine release in near real-time in vivo. However, contemporary FSCV techniques have been limited to measure only rapid (on the order of seconds, i.e. phasic) changes in dopamine release evoked by either electrical stimulation or elicited by presentation of behaviorally salient stimuli, and not slower changes in the tonic extracellular levels of dopamine (i.e. basal concentrations). This is because FSCV is inherently a differential method that requires subtraction of prestimulation tonic levels of dopamine to measure phasic changes relative to a zeroed baseline. Here, we describe the development and application of a novel voltammetric technique, multiple cyclic square wave voltammetry (M-CSWV), for analytical quantification of tonic dopamine concentrations in vivo with relatively high temporal resolution (10 s). M-CSWV enriches the electrochemical information by generating two dimensional voltammograms which enable high sensitivity (limit of detection, 0.17 nM) and selectivity against ascorbic acid, and 3,4-dihydroxyphenylacetic acid (DOPAC), including changes in pH. Using M-CSWV, a tonic dopamine concentration of 120 ± 18 nM (n = 7 rats, ± SEM) was determined in the striatum of urethane anethetized rats. Pharmacological treatments to elevate dopamine by selectively inhibiting dopamine reuptake and to reduce DOPAC by inhibition of monoamine oxidase supported the selective detection of dopamine in vivo. Overall, M-CSWV offers a novel voltammetric technique to quantify levels and monitor changes in tonic dopamine concentrations in the brain to further our understanding of the role of dopamine in normal behavior and neuropsychiatric disorders.

Fast Cyclic Square-Wave Voltammetry To Enhance Neurotransmitter Selectivity and Sensitivity

Although fast-scan cyclic voltammetry (FSCV) has been widely used for in vivo neurochemical detection, the sensitivity and selectivity of the technique can be further improved. In this study, we develop fast cyclic square-wave voltammetry (FCSWV) as a novel voltammetric technique that combines large-amplitude cyclic square-wave voltammetry (CSWV) with background subtraction. A large-amplitude, square-shaped potential was applied to induce cycling through multiple redox reactions within a square pulse to increase sensitivity and selectivity when combined with a two-dimensional voltammogram. As a result, FCSWV was significantly more sensitive than FSCV ( n = 5 electrodes, two-way ANOVA, p = 0.0002). In addition, FCSWV could differentiate dopamine from other catecholamines (e.g., epinephrine and norepinephrine) and serotonin better than conventional FSCV. With the confirmation that FCSWV did not influence local neuronal activity, despite the large amplitude of the square waveform, it could monitor electrically induced phasic changes in dopamine release in rat striatum before and after injecting nomifensine, a dopamine reuptake inhibitor.

Ultrafast charge and discharge biscrolled yarn supercapacitors for textiles and microdevices

Flexible, wearable, implantable and easily reconfigurable supercapacitors delivering high energy and power densities are needed for electronic devices. Here we demonstrate weavable, sewable, knottable and braidable yarns that function as high performance electrodes of redox supercapacitors. A novel technology, gradient biscrolling, provides fast-ion-transport yarn in which hundreds of layers of conducting-polymer-infiltrated carbon nanotube sheet are scrolled into ~20 μm diameter yarn. Plying the biscrolled yarn with a metal wire current collector increases power generation capabilities. The volumetric capacitance is high (up to ~179 F cm(-3)) and the discharge current of the plied yarn supercapacitor linearly increases with voltage scan rate up to ~80 V s(-1) and ~20 V s(-1) for liquid and solid electrolytes, respectively. The exceptionally high energy and power densities for the complete supercapacitor, and high cycle life that little depends on winding or sewing (92%, 99% after 10,000 cycles, respectively) are important for the applications in electronic textiles.