Home » Key Stem Cells Research Articles » Signal analysis and classification methods for the calcium transient data of stem cell-derived cardiomyocytes

Signal analysis and classification methods for the calcium transient data of stem cell-derived cardiomyocytes

Significance Statement

Calcium cycling has an important role for cardiomyocytes. It is the link between the electrical signaling in the cardiomyocyte and contraction. Changes and variability in calcium cycling are seen for the sake of cardiac diseases or drugs. Cardiac functionality can be investigated by means of cardiomyocytes differentiated from human pluripotent  stem cells. The technology of induced pluripotent stem cells (iPSC) gives a highly useful method for studying pathophysiology of disorders and drug responses. Human iPSCs can be differentiated into the desired cell type, retaining the genuine genotype. To empower cardiologic investigations, we have developed computational methods on the basis of signal analysis and classification with machine learning methods that aid to recognize peaks or cyclings (transients) in cardiomyocyte signal data and to classify entire signals into either a normal or abnormal class. The purpose is to create software tools for the selection of valid cell lines, observation of abnormal calcium cyclings and analysis of drug responses. Fig. 1 shows an example in which all signal peaks are recognized to be normal, i.e., the whole signal is normal. In Fig. 2 most of peaks are recognized to be abnormal. If even one peak only is seen abnormal, the entire signal is classified to be such. It is essential to develop automatic tools for the data analysis of cell lines creating really big data in order to aid biomedical researchers in their studies and, in the long term, to enable industrial utilization of novel discoveries in cell biology and medicine.

Figure Legend: Cardiomyocytes were exposed to two different wavelengths of light and emissions recorded. For calcium analysis, regions of interest were selected from a video stream of spontaneously beating cells. A signal (mean removed) of around 12 s with all peaks recognized normal represents a normal, valid calcium cycling waveform.

Signal analysis classification methods for ca+2 transient data of stem cell-derived cardiomyocytes. Global Medical Discovery

Figure Legend : A signal of peaks recognized to be abnormal except the last one represents an abnormal cycling waveform.

calcium transient data of stem cell-derived cardiomyocytes. Global Medical Discovery

 

 

 

 

 

 

 

 

 

 

 

Journal Reference

Comput Biol Med. 2015 Jun;61:1-7.

Juhola M1, Penttinen K2, Joutsijoki H3, Varpa K3, Saarikoski J3, Rasku J3, Siirtola H3, Iltanen K3, Laurikkala J3, Hyyrö H3, Hyttinen J4, Aalto-Setälä K5.

Show Affiliations

1School of Information Sciences, University of Tampere, Tampere, Finland. Electronic address: [email protected]

2BioMediTech and School of Medicine, University of Tampere, Tampere, Finland.

3School of Information Sciences, University of Tampere, Tampere, Finland.

4Department of Electronics and Communication Engineering, Tampere University of Technology, Tampere, Finland.

5BioMediTech and School of Medicine, University of Tampere, Tampere, Finland; Heart Hospital, Tampere University Hospital, Tampere, Finland.

Abstract

Calcium cycling is crucial in the excitation-contraction coupling of cardiomyocytes, and therefore has a key role in cardiac functionality. Cardiac disorders and different drugs alter the calcium transients of cardiomyocytes and can cause serious dysfunction of the heart. New insights into this biochemical phenomena can be achieved by studying and analyzing calcium transients. Calcium transients of spontaneously beating human induced pluripotent stem cell-derived cardiomyocytes were recorded for a data set of 280 signals. Our objective was to develop and program procedures: (1) to automatically detect cycling peaks from signals and to classify the peaks of signals as either normal or abnormal, and (2) on the basis of the preceding peak detection results, to classify the entire signals into either a normal class or an abnormal class. We obtained a classification accuracy of approximately 80% compared to class decisions made separately by an experienced researcher, which is promising for the further development of an automatic classification approach. Automated classification software would be beneficial in the future for analyzing cardiomyocyte functionality on a large scale when screening for the adverse cardiac effects of new potential compounds, and also in future clinical applications.

Copyright © 2015 Elsevier Ltd. All rights reserved.

Go To Comput Biol Med.