DEVELOPMENT OF A SEMI-AUTOMATIC ALGORITHM FOR DECONVOLUTION AND QUANTIFICATION OF THREE-DIMENSIONAL MICROSCOPY IMAGES

Modern microscopy enables the acquisition of massive volumes of information. Processing and evaluating large multidimensional images is time-consuming, especially when working with various stacks. In the present work we have developed a software tool for the optimization of image processing, which consists of an automatic deconvolution and quantification algorithm that eliminates non-systematic errors and significantly decreases processing time. This tool included a restoration deconvolution method (positive constrained algorithm) and five image-restoration parameters (Contrast-to-Noise Ratio, Signal-to-Noise Ratio, Full-Width at Half-Maximum and two three-dimensional Tenengrad-based indicators) used to assess quantitatively the quality of restoration. This algorithm was used to process raw three-dimensional images using several experimental Point Spread Functions; raw images were obtained by fluorescence wide-field microscopy of epidermal E-cadherin expression in Rhinella (= Bufo) arenarum embryos and fluorescent microspheres. The image-restoration indicators and the performance of the algorithm were evaluated. Results show that all indicators concur and do not increase processing time significantly, constituting a valuable tool for 3D microscopy analysis.