Accurate detection of low signal-to-noise ratio neuronal calcium transient waves using a matched filter

Agnieszka F. Szymanska, Chiaki Kobayashi, Hiroaki Norimoto, Tomoe Ishikawa, Yuji Ikegaya, Zoran Nenadic

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Background: Calcium imaging has become a fundamental modality for studying neuronal circuit dynamics both in vitro and in vivo. However, identifying calcium events (CEs) from spectral data remains laborious and difficult, especially since the signal-to-noise ratio (SNR) often falls below 2. Existing automated signal detection methods are generally applied at high SNRs, leaving a large need for an automated algorithm that can accurately extract CEs from fluorescence intensity data of SNR 2 and below. New method: In this work we develop a Matched filter for Mult. i-unit Calcium Event (MMiCE) detection to extract CEs from fluorescence intensity traces of simulated and experimentally recorded neuronal calcium imaging data. Results: MMiCE reached perfect performance on simulated data with SNR ≥ 2 and a true positive (TP) rate of 98.27% (± 1.38% with a 95% confidence interval), and a false positive(FP) rate of 6.59% (± 2.56%) on simulated data with SNR 0.2. On real data, verified by patch-clamp recording, MMiCE performed with a TP rate of 100.00% (± 0.00) and a FP rate of 2.04% (± 4.10). Comparison with existing method(s): This high level of performance exceeds existing methods at SNRs as low as 0.2, which are well below those used in previous studies (SNR ≃ 5-10). Conclusion: Overall, the MMiCE detector performed exceptionally well on both simulated data, and experimentally recorded neuronal calcium imaging data. The MMiCE detector is accurate, reliable, well suited for wide-spread use, and freely available at sites.uci.edu/aggies or from the corresponding author.

Original languageEnglish
Pages (from-to)1-12
Number of pages12
JournalJournal of Neuroscience Methods
Volume259
DOIs
Publication statusPublished - 2016 Feb 1
Externally publishedYes

Fingerprint

Calcium Signaling
Signal-To-Noise Ratio
Calcium
Fluorescence
Confidence Intervals

Keywords

  • Calcium transients
  • Dendritic spines
  • Detection
  • Low SNR
  • Matched filter
  • Multineuron calcium imaging
  • Somatic calcium fluctuations

ASJC Scopus subject areas

  • Neuroscience(all)

Cite this

Accurate detection of low signal-to-noise ratio neuronal calcium transient waves using a matched filter. / Szymanska, Agnieszka F.; Kobayashi, Chiaki; Norimoto, Hiroaki; Ishikawa, Tomoe; Ikegaya, Yuji; Nenadic, Zoran.

In: Journal of Neuroscience Methods, Vol. 259, 01.02.2016, p. 1-12.

Research output: Contribution to journalArticle

Szymanska, Agnieszka F. ; Kobayashi, Chiaki ; Norimoto, Hiroaki ; Ishikawa, Tomoe ; Ikegaya, Yuji ; Nenadic, Zoran. / Accurate detection of low signal-to-noise ratio neuronal calcium transient waves using a matched filter. In: Journal of Neuroscience Methods. 2016 ; Vol. 259. pp. 1-12.
@article{324cd9cc6c0d49b1ae2aa8626bed9eb9,
title = "Accurate detection of low signal-to-noise ratio neuronal calcium transient waves using a matched filter",
abstract = "Background: Calcium imaging has become a fundamental modality for studying neuronal circuit dynamics both in vitro and in vivo. However, identifying calcium events (CEs) from spectral data remains laborious and difficult, especially since the signal-to-noise ratio (SNR) often falls below 2. Existing automated signal detection methods are generally applied at high SNRs, leaving a large need for an automated algorithm that can accurately extract CEs from fluorescence intensity data of SNR 2 and below. New method: In this work we develop a Matched filter for Mult. i-unit Calcium Event (MMiCE) detection to extract CEs from fluorescence intensity traces of simulated and experimentally recorded neuronal calcium imaging data. Results: MMiCE reached perfect performance on simulated data with SNR ≥ 2 and a true positive (TP) rate of 98.27{\%} (± 1.38{\%} with a 95{\%} confidence interval), and a false positive(FP) rate of 6.59{\%} (± 2.56{\%}) on simulated data with SNR 0.2. On real data, verified by patch-clamp recording, MMiCE performed with a TP rate of 100.00{\%} (± 0.00) and a FP rate of 2.04{\%} (± 4.10). Comparison with existing method(s): This high level of performance exceeds existing methods at SNRs as low as 0.2, which are well below those used in previous studies (SNR ≃ 5-10). Conclusion: Overall, the MMiCE detector performed exceptionally well on both simulated data, and experimentally recorded neuronal calcium imaging data. The MMiCE detector is accurate, reliable, well suited for wide-spread use, and freely available at sites.uci.edu/aggies or from the corresponding author.",
keywords = "Calcium transients, Dendritic spines, Detection, Low SNR, Matched filter, Multineuron calcium imaging, Somatic calcium fluctuations",
author = "Szymanska, {Agnieszka F.} and Chiaki Kobayashi and Hiroaki Norimoto and Tomoe Ishikawa and Yuji Ikegaya and Zoran Nenadic",
year = "2016",
month = "2",
day = "1",
doi = "10.1016/j.jneumeth.2015.10.014",
language = "English",
volume = "259",
pages = "1--12",
journal = "Journal of Neuroscience Methods",
issn = "0165-0270",
publisher = "Elsevier",

}

TY - JOUR

T1 - Accurate detection of low signal-to-noise ratio neuronal calcium transient waves using a matched filter

AU - Szymanska, Agnieszka F.

AU - Kobayashi, Chiaki

AU - Norimoto, Hiroaki

AU - Ishikawa, Tomoe

AU - Ikegaya, Yuji

AU - Nenadic, Zoran

PY - 2016/2/1

Y1 - 2016/2/1

N2 - Background: Calcium imaging has become a fundamental modality for studying neuronal circuit dynamics both in vitro and in vivo. However, identifying calcium events (CEs) from spectral data remains laborious and difficult, especially since the signal-to-noise ratio (SNR) often falls below 2. Existing automated signal detection methods are generally applied at high SNRs, leaving a large need for an automated algorithm that can accurately extract CEs from fluorescence intensity data of SNR 2 and below. New method: In this work we develop a Matched filter for Mult. i-unit Calcium Event (MMiCE) detection to extract CEs from fluorescence intensity traces of simulated and experimentally recorded neuronal calcium imaging data. Results: MMiCE reached perfect performance on simulated data with SNR ≥ 2 and a true positive (TP) rate of 98.27% (± 1.38% with a 95% confidence interval), and a false positive(FP) rate of 6.59% (± 2.56%) on simulated data with SNR 0.2. On real data, verified by patch-clamp recording, MMiCE performed with a TP rate of 100.00% (± 0.00) and a FP rate of 2.04% (± 4.10). Comparison with existing method(s): This high level of performance exceeds existing methods at SNRs as low as 0.2, which are well below those used in previous studies (SNR ≃ 5-10). Conclusion: Overall, the MMiCE detector performed exceptionally well on both simulated data, and experimentally recorded neuronal calcium imaging data. The MMiCE detector is accurate, reliable, well suited for wide-spread use, and freely available at sites.uci.edu/aggies or from the corresponding author.

AB - Background: Calcium imaging has become a fundamental modality for studying neuronal circuit dynamics both in vitro and in vivo. However, identifying calcium events (CEs) from spectral data remains laborious and difficult, especially since the signal-to-noise ratio (SNR) often falls below 2. Existing automated signal detection methods are generally applied at high SNRs, leaving a large need for an automated algorithm that can accurately extract CEs from fluorescence intensity data of SNR 2 and below. New method: In this work we develop a Matched filter for Mult. i-unit Calcium Event (MMiCE) detection to extract CEs from fluorescence intensity traces of simulated and experimentally recorded neuronal calcium imaging data. Results: MMiCE reached perfect performance on simulated data with SNR ≥ 2 and a true positive (TP) rate of 98.27% (± 1.38% with a 95% confidence interval), and a false positive(FP) rate of 6.59% (± 2.56%) on simulated data with SNR 0.2. On real data, verified by patch-clamp recording, MMiCE performed with a TP rate of 100.00% (± 0.00) and a FP rate of 2.04% (± 4.10). Comparison with existing method(s): This high level of performance exceeds existing methods at SNRs as low as 0.2, which are well below those used in previous studies (SNR ≃ 5-10). Conclusion: Overall, the MMiCE detector performed exceptionally well on both simulated data, and experimentally recorded neuronal calcium imaging data. The MMiCE detector is accurate, reliable, well suited for wide-spread use, and freely available at sites.uci.edu/aggies or from the corresponding author.

KW - Calcium transients

KW - Dendritic spines

KW - Detection

KW - Low SNR

KW - Matched filter

KW - Multineuron calcium imaging

KW - Somatic calcium fluctuations

UR - http://www.scopus.com/inward/record.url?scp=84961894536&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84961894536&partnerID=8YFLogxK

U2 - 10.1016/j.jneumeth.2015.10.014

DO - 10.1016/j.jneumeth.2015.10.014

M3 - Article

C2 - 26561771

AN - SCOPUS:84961894536

VL - 259

SP - 1

EP - 12

JO - Journal of Neuroscience Methods

JF - Journal of Neuroscience Methods

SN - 0165-0270

ER -