Presentation Details
| Don Tucker Finalist: Distinguishing the Olfactory Epithelium using an FDA-Approved Dye & Machine Learning Methods Skylar A Suarez1, Emily A Gibson1, Diego Restrepo2, 3. 1Department of Bioengineering, University of Colorado Anschutz Medical Campus, Denver, CO, USA.2Department of Physiology & Biophysics, Stony Brook University, Stony Brook, NY, USA.3Department of Cell & Developmental Biology, University of Colorado Anschutz Medical Campus, Denver, CO, USA |
Abstract
The olfactory epithelium (OE) is a specialized tissue located deep within the nasal cavity that contains stem cells (globose basal cells and horizontal basal cells), supporting cells and olfactory sensory neurons (OSNs) that are responsible for detecting odorants and transmitting olfactory information to the brain. Anosmia is common in aging and can be a symptom of respiratory infections, medication side effects, and some neurological disorders. As humans age, some OE areas undergo neurogenic exhaustion as their OSNs stop being replaced by globose basal cells, leading to patchiness. In other conditions, the OE may be damaged. As a result, the OE exists in patches surrounded by exhausted OE and respiratory epithelia (RE) within the nasal cavity. These patches vary in size and the OSNs are not uniformly distributed throughout, especially in aging and conditions related to olfactory dysfunction. Therefore, the ability to visualize the OE in vivo can assist with assessment of the cause of aging-related anosmia and olfactory dysfunction. It can also assist in observing and quantifying any changes in OE associated with progression or treatments. This project explores using Indocyanine Green to visualize nasal epithelia tissue structure, followed machine learning and deep learning methods to distinguish the OE from its surrounding respiratory epithelia. Using texture analysis, the machine learning methods reached ~80% classification accuracy, while direct deep learning methods reached ~90% accurate classification of OE images.
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.