Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (7 results)

  • Article
  • Peer Reviewed
UC Irvine Previously Published Works (2014)
Linear approaches to low-dimensional feature extraction may not be appropriate when statistical data are generated by a nonlinear interaction of parameters. Equally inadequate are linear methods for determining the dimension of the feature space. This article estimates the intrinsic dimension of extracellular action potentials (EAPs), which can be viewed as the minimum number of nonlinearly interacting parameters sufficient to describe the data. When combined with nonlinear feature extraction methods, this information may lead to a more faithful, low-dimensional EAP representation. These points are demonstrated using EAPs recorded experimentally by a multisensor electrode.
    Cover page: Intrinsic Dimensionality of Extracellular Action PotentialsCreative Commons 'BY' version 4.0 license
    • Thesis
    • Peer Reviewed
    UC Irvine Electronic Theses and Dissertations (2017)

    Electrophysiological signal processing is a broad, complex and growing field. Universally, the first steps of any signal analysis are detection and classification. Here we present a flexible matched filter designed to detect spikes from various biological data types, as well as two statistically based approaches for spike classification. We then apply these developed tools to study the effects of deep anesthesia on neuronal network dynamics.

    The matched filter was implemented for three different applications: detecting action potentials (APs) from multi-sensor extracellular recordings, detecting depolarization events (DEs) from voltage sensitive dye (VSD) imaged cardiomyocytes, and detecting calcium events (CEs) from calcium imaged neuronal somas as well as dendritic spines. Overall, the presented matched filters could accurately detect spikes from various kinds of biological data, often beating other existing methods, and outperforming manual spike selection.

    The classification problems explored here include AP (or spike) sorting, as well as DE classification across different drug administrations. In the case of spike sorting, the MUSIC algorithm was used to extract classification features from multi-sensor extracellular AP recordings. This approach was able to reliably classify tetrode (4 channel) and heptode (7 channel) recorded APs. For DE classification, salient DE features were extracted and then compared across drug treatments using a Kolmogorov-Smirnov test. The drug treated cells were consistently statistically distinguishable from controls. Overall, both methods’ success makes them valuable tools for studying neuronal networks as well as cardiomyocyte drug assays, respectively.

    Finally, the matched filter for AP detection as well as the MUSIC-based AP classification scheme were applied to in vivo heptode data collected from M1 of the right hemisphere of anesthetized rats, to assess the effects of increasing anesthesia on neuronal network dynamics during a burst suppression state. We found that higher anesthesia led to higher AP frequency, no change in the number of active single units, and increased cross-hemisphere functional connectivity (as measured by simultaneous ECoG recordings). Additionally, all APs were restricted to ECoG bursts, with no APs occurring during suppressed ECoG states. This work provides valuable insights for the study of neuronal network dynamics as well as coma arousal.

      Cover page: Action Potentials and Waves: A Short Story on Electrophysiological Signal ProcessingCreative Commons 'BY-ND' version 4.0 license
      • Article
      • Peer Reviewed
      UC Irvine Previously Published Works (2013)
      Four-sensor microelectrodes, commonly referred to as tetrodes, have the ability to significantly increase the signal-to-noise ratio of neuronal extracellular recordings. They also provide spatio-temporal information about extracellular action potentials (EAP) which may be used to localize and resolve individual neuronal signal sources. Since the relative position of sensors and neurons whose EAPs are recorded is not known during in vivo experiments, the accuracy and precision of neuronal source localization algorithms remain untested. In this study, electrical signals generated by a stimulator were recorded simultaneously with four recording micropipettes immersed in artificial cerebrospinal fluid. The location of the source was estimated using the multiple signal classification algorithm, with an accuracy and precision of ~4 µm and ~7 µm, respectively. These results suggest that in vivo localization and resolution of individual neuronal sources is feasible.
        Cover page: The Accuracy and Precision of Signal Source Localization with TetrodesCreative Commons 'BY' version 4.0 license
        • Article
        • Peer Reviewed
        UC Irvine Previously Published Works (2014)
        Multi-sensor extracellular recording takes advantage of several electrode channels to record from multiple neurons at the same time. However, the resulting low signal-to-noise ratio (SNR) combined with biological noise makes signal detection, the first step of any neurophysiological data analysis, difficult. A matched filter was therefore designed to better detect extracellular action potentials (EAPs) from multi-sensor extracellular recordings. The detector was tested on tetrode data from a locust antennal lobe and assessed against three trained analysts. 25 EAPs and noise samples were selected manually from the data and used for training. To reduce complexity, the filter assumed that the underlying noise in the data was spatially white. The detector performed with an average TP and FP rate of 84.62% and 16.63% respectively. This high level of performance indicates the algorithm is suitable for widespread use.
          Cover page: A Supervised Multi-Sensor Matched Filter for the Detection of Extracellular Action PotentialsCreative Commons 'BY' version 4.0 license
          • Article
          • Peer Reviewed
          UC Irvine Previously Published Works (2013)
          Multi-sensor microelectrodes for extracellular action potential recording have significantly improved the quality of in vivo recorded neuronal signals. These microelectrodes have also been instrumental in the localization of neuronal signal sources. However, existing neuron localization methods have been mostly utilized in vivo, where the true neuron location remains unknown. Therefore, these methods could not be experimentally validated. This article presents experimental validation of a method capable of estimating both the location and intensity of an electrical signal source. A four-sensor microelectrode (tetrode) immersed in a saline solution was used to record stimulus patterns at multiple intensity levels generated by a stimulating electrode. The location of the tetrode was varied with respect to the stimulator. The location and intensity of the stimulator were estimated using the Multiple Signal Classification (MUSIC) algorithm, and the results were quantified by comparison to the true values. The localization results, with an accuracy and precision of ~ 10 microns, and ~ 11 microns respectively, imply that MUSIC can resolve individual neuronal sources. Similarly, source intensity estimations indicate that this approach can track changes in signal amplitude over time. Together, these results suggest that MUSIC can be used to characterize neuronal signal sources in vivo.
            Cover page: A Method for Neuronal Source IdentificationCreative Commons 'BY' version 4.0 license
            • Article
            • Peer Reviewed
            UC Irvine Previously Published Works (2016)
            Optical imaging-based methods for assessing the membrane electrophysiology of in vitro human cardiac cells allow for non-invasive temporal assessment of the effect of drugs and other stimuli. Automated methods for detecting and analyzing the depolarization events (DEs) in image-based data allow quantitative assessment of these different treatments. In this study, we use 2-photon microscopy of fluorescent voltage-sensitive dyes (VSDs) to capture the membrane voltage of actively beating human induced pluripotent stem cell-derived cardiomyocytes (hiPS-CMs). We built a custom and freely available Matlab software, called MaDEC, to detect, quantify, and compare DEs of hiPS-CMs treated with the β-adrenergic drugs, propranolol and isoproterenol. The efficacy of our software is quantified by comparing detection results against manual DE detection by expert analysts, and comparing DE analysis results to known drug-induced electrophysiological effects. The software accurately detected DEs with true positive rates of 98-100% and false positive rates of 1-2%, at signal-to-noise ratios (SNRs) of 5 and above. The MaDEC software was also able to distinguish control DEs from drug-treated DEs both immediately as well as 10min after drug administration.
              Cover page: Automated detection and analysis of depolarization events in human cardiomyocytes using MaDEC.
              • Article
              • Peer Reviewed
              UC Irvine Previously Published Works (2013)
              Electrocorticogram (ECoG)-based brain computer interfaces (BCI) can potentially be used for control of arm prostheses. Restoring independent function to BCI users with such a system will likely require control of many degrees-of-freedom (DOF). However, our ability to decode many-DOF arm movements from ECoG signals has not been thoroughly tested. To this end, we conducted a comprehensive study of the ECoG signals underlying 6 elementary upper extremity movements. Two subjects undergoing ECoG electrode grid implantation for epilepsy surgery evaluation participated in the study. For each task, their data were analyzed to design a decoding model to classify ECoG as idling or movement. The decoding models were found to be highly sensitive in detecting movement, but not specific in distinguishing between different movement types. Since sensitivity and specificity must be traded-off, these results imply that conventional ECoG grids may not provide sufficient resolution for decoding many-DOF upper extremity movements.
                Cover page: Sensitivity and specificity of upper extremity movements decoded from electrocorticogram.Creative Commons 'BY' version 4.0 license
                Top