Presentation Details
| Automated 3D Tracking of Taste Bud Cells for Morphological & Lifespan Analysis Brittany N.Walters, David C.Alston, Robin F.Krimm, . University of Louisville School of Medicine, Louisville, KY, USA |
Abstract
Taste bud cells continually die and are replaced, necessitating methods to study their dynamics. This project uses in vivo dual channel two-photon microscopy, performed in mice, to collect the same taste buds over time. Here, we present an analytic pipeline for 3D automated segmentation and tracking of individual taste bud cells over time. Due to the densely packed nature of cells in taste buds, manual cell segmentation and tracking are labor-intensive and inefficient. Automating these processes allows for more efficient analysis of cell morphology, spatial position, and lifespan, significantly increasing the volume of data that can be analyzed. Napari, a Python-based tool for image processing, is used to render volumetric reconstructions and generate ground-truth segmentations of individual cells. For preprocessing, we used Noise2Void, and for registration of taste bud volumes collected at successive time points, we used the Correct 3D Drift plugin in ImageJ. Concerning automated segmentation, we used a U-Net that uses transformers (SeUNet), while Ultrack tracks segmentations of individual cells over time. We have established proof of concept demonstrating that the affinity-based segmentation model performs well in sparsely populated taste buds. Based on this, we are refining the affinity model to improve performance in densely packed taste buds. Cell morphological measurements will be quantified using the MorphoLibJ plugin in ImageJ. To our knowledge, this is the first integrated tool specifically designed for automated 3D segmentation and longitudinal tracking of individual taste bud cells. It enables analysis of cell morphology and spatial position within the taste bud, measures total cell lifespan, and provides a framework for other 3D time-lapse imaging studies.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
Content Locked. Log into a registered attendee account to access this presentation.