Combining Morphological and Textural Features to Characterize Mitochondrial Structure

Abstract

Understanding cellular microenvironments and cellular responses is crucial when evaluating treatments and/or pathological conditions. Changes in sub-cellular structures can be assessed by imaging and used to evaluate cellular responses to external factors. One of the key sub-cellular structures that can be assessed by imaging and used as cues to correlate cellular responses with external factors. One of the key sub-cellular structures which can provide us with information about cellular metabolic states are mitochondria. They are the cell ‘powerhouses’ responsible for energy production and produce bioactive molecules that regulate other cellular functions. Thus, investigating changes in mitochondria organization/structure and their localization within the cell can provide important insights into cell-microenvironment interactions. The mitochondrial structure changes dynamically and the main formed structures can be classified as network-like arrangements, puncti, or rods. In this study, super-resolution confocal microscopy was used to assess mitochondrial structure and a combination of image-based morphological and textural features are used to extractive objective information about the main mitochondrial structure’s shapes such as fiber orientation, length, concentration based on pixel intensities, degree of structural organization and other related features. Our preliminary results show that morphological feature-based metrics were not successful in differentiating the mitochondrial arrangements. However, textural features provided the potential to allow the quantification of differences between puncta and fiber-network structures. Rods-like structures are still under investigation. By combining morphological and textural features we aim to create an image-based score which will account for all three main mitochondrial structures commonly present in cells and then allow us to perform a systematic image-based differentiation of cellular metabolic states.