Abstract
Plant cell contains a number of organelles. Cellular activities are supported by the cooperative functions of organelles. Analysis of microscopic images play an essential role in organelle research. Specially, recent technical advance of live imaging allow us to obtain large data sets of time-sequential images of each organelle. However, in contrast to biosequential data and gene expression data, quantitative analysis are rarely performed for organelle images because of the diversity of imaging-technique including purposes, targets and conditions. To develop a versatile technique for organelle image analysis, we have focused on endoplasmic reticulum (ER) flow as a model case of organelle dynamics and investigated the method of speed measurement from time-sequential images. As a result, we succeeded to develop a software tool for flow analysis using local and spatio-temporal image correlation spectrum. Furthermore, we have investigated to measure the movements for multiple organelles visualized in single fluorescence band. As part of the investigation, we are currently attempting to develop the machine learning based image classifier and distribution estimator of organelles.
