Developing analysis methods for signals and images arising from imaging calcium experiments >>> Why developing new analysis methods for calcium imaging? Calcium imaging has been used for more than 20 years to study both cellular and networks properties of living tissues, and is popular for its allowance to record simultaneously the activity of several neurons. My post-doc is in line with a collaboration with Peter Kloppenburg's lab in Cologne (Germany). One of their projects aims at characterizing the endogenous binding capacity of interneurons in the olfactory bulb of an insect, the Periplaneta Americana cockroach. For that purpose, they use the 'added buffer approach', which consists in following the calcium dynamics during the intracellular loading of an exogenous calcium buffer (for instance, Fura-2). During the loading procedure, several electrical stimulations are delivered to the cell, leading to variations of the intracellular free calcium concentration. From the estimation of the time constants of these variations, it is then possible to estimate the endogenous binding capacity of the target cell. >>> To get a reliable estimation of this binding capacity, two main steps must be achieved: 1- detect the region(s) of interest in the fluorescence images sequence. 2- estimate in a quantitative way the parameters of the calcium dynamics extracted from the images sequence. >>> To date, the second point has been mainly adressed. We focused on the development of a data generation model for fluorescence data obtained with a ratiometric dye and a CCD camera. A key feature of this model is the consideration of the characteristics of the acquisition system, which soils the measurement with a Poissonian noise. Taking into account this key property, we developed a direct method to estimate the calcium dynamics, without using the classical ratiometric transformation. Compared to the latter, our method has three main advantages: 1) Confidence intervals provided for parametric models of calcium dynamics are reliable. 2) The uncertainty on the calibration parameters are taken into account. 3)
Moreover, we can estimate the variations of the dye intracellular concentration throughout the experiment, which leads to improved estimations of the calcium dynamics. This new method has been presented last November at the Annual Meeting of the Society for Neuroscience (see the poster). It has been implemented in CalciOMatic, a new R package, which will be soon available from this site. >>> In a near future, we will focus on developing a method to select automatically the regions of interest. |
|
|
|
Post-doc |