Advanced Techniques in UV Microscopy: Integrating Deep Learning for Autofocusing, Whole Slide Imaging Optimization, and Spicule Detection in Bone Marrow Aspirations
Loading...
Author(s)
Subramanian, Ajay Rajaraman
Advisor(s)
Editor(s)
Collections
Supplementary to:
Permanent Link
Abstract
Deep-ultraviolet (UV) microscopy enables label-free, high-resolution, quantitative molecular imaging and enables unique applications in biomedicine. This is achieved by
leveraging the unique absorption of different biomolecules in the deep UV spectrum (200~400nm). While UV microscopy has historically been limited by phototoxicity concerns, recent advances in UV imaging hardware have allowed for live cell imaging for more than six hours with no observable photodamage. This proposal aims to expand the capabilities of UV microscopy for biomedical applications using CNN based architectures. Initially, we will employ a custom benchtop
multispectral microscope for single-shot autofocusing. Subsequently, we will utilize a lowcost compact UV microscope for automated whole-slide imaging, incorporating singleshot autofocusing as well. Additionally, the compact UV microscope will be employed for automated spicule detection in bone marrow aspirations to enable an untrained clinician to
use the device.
Sponsor
Date
2024-05-02
Extent
Resource Type
Text
Resource Subtype
Thesis