A new approach to the detection and statistical classification of Ca 2+ sparks

Tamás Bányász, Ye Chen-Izu, C. W. Balke, Leighton T Izu

Research output: Contribution to journalArticle

26 Citations (Scopus)

Abstract

The availability of high-speed, two-dimensional (2-D) confocal microscopes and the expanding armamentarium of fluorescent probes presents unprecedented opportunities and new challenges for studying the spatial and temporal dynamics of cellular processes. The need to remove subjectivity from the detection process, the difficulty of the human eye to detect subtle changes in fluorescence in these 2-D images, and the large volume of data produced by these confocal microscopes call for the need to develop algorithms to automatically mark the changes in fluorescence. These fluorescence signal changes are often subtle, so the statistical estimate of the likelihood that the detected signal is not noise is an integral part of the detection algorithm. This statistical estimation is fundamental to our new approach to detection; in earlier Ca 2+ spark detectors, this statistical assessment was incidental to detection. Importantly, the use of the statistical properties of the signal local to the spark, instead of over the whole image, reduces the false positive and false negative rates. We developed an automatic spark detection algorithm based on these principles and used it to detect sparks on an inhomogeneous background of transverse tubule-labeled rat ventricular cells. Because of the large region of the cell surveyed by the confocal microscope, we can detect a large enough number of sparks to measure the dynamic changes in spark frequency in individual cells. We also found, in contrast to earlier results, that cardiac sparks are spatially symmetric. This new approach puts the detection of fluorescent signals on a firm statistical foundation.

Original languageEnglish (US)
Pages (from-to)4458-4465
Number of pages8
JournalBiophysical Journal
Volume92
Issue number12
DOIs
StatePublished - Jun 2007
Externally publishedYes

Fingerprint

Fluorescence
Fluorescent Dyes
Noise

ASJC Scopus subject areas

  • Biophysics

Cite this

A new approach to the detection and statistical classification of Ca 2+ sparks. / Bányász, Tamás; Chen-Izu, Ye; Balke, C. W.; Izu, Leighton T.

In: Biophysical Journal, Vol. 92, No. 12, 06.2007, p. 4458-4465.

Research output: Contribution to journalArticle

@article{4065d75fa0534b4e8cbc1e6917a9a368,
title = "A new approach to the detection and statistical classification of Ca 2+ sparks",
abstract = "The availability of high-speed, two-dimensional (2-D) confocal microscopes and the expanding armamentarium of fluorescent probes presents unprecedented opportunities and new challenges for studying the spatial and temporal dynamics of cellular processes. The need to remove subjectivity from the detection process, the difficulty of the human eye to detect subtle changes in fluorescence in these 2-D images, and the large volume of data produced by these confocal microscopes call for the need to develop algorithms to automatically mark the changes in fluorescence. These fluorescence signal changes are often subtle, so the statistical estimate of the likelihood that the detected signal is not noise is an integral part of the detection algorithm. This statistical estimation is fundamental to our new approach to detection; in earlier Ca 2+ spark detectors, this statistical assessment was incidental to detection. Importantly, the use of the statistical properties of the signal local to the spark, instead of over the whole image, reduces the false positive and false negative rates. We developed an automatic spark detection algorithm based on these principles and used it to detect sparks on an inhomogeneous background of transverse tubule-labeled rat ventricular cells. Because of the large region of the cell surveyed by the confocal microscope, we can detect a large enough number of sparks to measure the dynamic changes in spark frequency in individual cells. We also found, in contrast to earlier results, that cardiac sparks are spatially symmetric. This new approach puts the detection of fluorescent signals on a firm statistical foundation.",
author = "Tam{\'a}s B{\'a}ny{\'a}sz and Ye Chen-Izu and Balke, {C. W.} and Izu, {Leighton T}",
year = "2007",
month = "6",
doi = "10.1529/biophysj.106.103069",
language = "English (US)",
volume = "92",
pages = "4458--4465",
journal = "Biophysical Journal",
issn = "0006-3495",
publisher = "Biophysical Society",
number = "12",

}

TY - JOUR

T1 - A new approach to the detection and statistical classification of Ca 2+ sparks

AU - Bányász, Tamás

AU - Chen-Izu, Ye

AU - Balke, C. W.

AU - Izu, Leighton T

PY - 2007/6

Y1 - 2007/6

N2 - The availability of high-speed, two-dimensional (2-D) confocal microscopes and the expanding armamentarium of fluorescent probes presents unprecedented opportunities and new challenges for studying the spatial and temporal dynamics of cellular processes. The need to remove subjectivity from the detection process, the difficulty of the human eye to detect subtle changes in fluorescence in these 2-D images, and the large volume of data produced by these confocal microscopes call for the need to develop algorithms to automatically mark the changes in fluorescence. These fluorescence signal changes are often subtle, so the statistical estimate of the likelihood that the detected signal is not noise is an integral part of the detection algorithm. This statistical estimation is fundamental to our new approach to detection; in earlier Ca 2+ spark detectors, this statistical assessment was incidental to detection. Importantly, the use of the statistical properties of the signal local to the spark, instead of over the whole image, reduces the false positive and false negative rates. We developed an automatic spark detection algorithm based on these principles and used it to detect sparks on an inhomogeneous background of transverse tubule-labeled rat ventricular cells. Because of the large region of the cell surveyed by the confocal microscope, we can detect a large enough number of sparks to measure the dynamic changes in spark frequency in individual cells. We also found, in contrast to earlier results, that cardiac sparks are spatially symmetric. This new approach puts the detection of fluorescent signals on a firm statistical foundation.

AB - The availability of high-speed, two-dimensional (2-D) confocal microscopes and the expanding armamentarium of fluorescent probes presents unprecedented opportunities and new challenges for studying the spatial and temporal dynamics of cellular processes. The need to remove subjectivity from the detection process, the difficulty of the human eye to detect subtle changes in fluorescence in these 2-D images, and the large volume of data produced by these confocal microscopes call for the need to develop algorithms to automatically mark the changes in fluorescence. These fluorescence signal changes are often subtle, so the statistical estimate of the likelihood that the detected signal is not noise is an integral part of the detection algorithm. This statistical estimation is fundamental to our new approach to detection; in earlier Ca 2+ spark detectors, this statistical assessment was incidental to detection. Importantly, the use of the statistical properties of the signal local to the spark, instead of over the whole image, reduces the false positive and false negative rates. We developed an automatic spark detection algorithm based on these principles and used it to detect sparks on an inhomogeneous background of transverse tubule-labeled rat ventricular cells. Because of the large region of the cell surveyed by the confocal microscope, we can detect a large enough number of sparks to measure the dynamic changes in spark frequency in individual cells. We also found, in contrast to earlier results, that cardiac sparks are spatially symmetric. This new approach puts the detection of fluorescent signals on a firm statistical foundation.

UR - http://www.scopus.com/inward/record.url?scp=34250307211&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34250307211&partnerID=8YFLogxK

U2 - 10.1529/biophysj.106.103069

DO - 10.1529/biophysj.106.103069

M3 - Article

C2 - 17400702

AN - SCOPUS:34250307211

VL - 92

SP - 4458

EP - 4465

JO - Biophysical Journal

JF - Biophysical Journal

SN - 0006-3495

IS - 12

ER -