On the development of non-invasive diagnosis tools with optical wavefront shaping
Loading...
Authors
Rumman, Nazifa
Issue Date
2024-05
Type
Electronic thesis
Thesis
Thesis
Language
en_US
Keywords
Electrical engineering
Alternative Title
Abstract
The use of light has reformed our approaches to diagnose or treat diseases, and transformed our understanding of physiological processes. Many of today’s optical technologies such as fluorescence microscopy, optogenetics, or photodynamic therapy rely on the precise delivery of light to or from target cells or tissues. However, the scattering of light as it penetrates through skin or tissue poses a significant challenge, restricting the performance of these techniques, limiting imaging depth, and requiring invasive procedures. A plausible solution to this problem lies in the ability to control the scattered light in a desired manner, which has been made feasible by wavefront shaping methods. Wavefront Shaping methods hold the potential to overcome the limitations of these optical technologies by increasing imaging depth, enhancing signal and sensitivity.Wavefront shaping methods can be applied to focus light through scattering media at a predefined target. However, in order to utilize the full potential of the biological applications such as deep tissue fluorescence imaging, we need to extend beyond generating a single focus. We have to consider signal extraction from targets hidden behind scattering media, generation of multiple focus points, and simultaneous enhancement of signals emitted by multiple fluorescent targets.
Here, we present an optical wavefront shaping method that allows enhancement of ini- tially weak signal emitted by sources (i.e., fluorescent particles) hidden behind a scattering medium. We propose a method that combines traditional feedback based wavefront shaping method together with a switch function and demonstrates utilization of two different sig- nals for shaping the incident wavefront. By taking advantage of the instantaneous changes induced by the modulation of the incident light, the hidden targets are located at first. A switch function introduced in the algorithm then advances optimization of the hidden sig- nal. Additionally, we present a simple, easy to implement scoring based genetic algorithm (SBGA) that can obtain multiple objectives simultaneously. Each possible solution acquires scores based on its capability to optimize individual objectives and the scores are combined to identify the optimal solution. We experimentally demonstrate our method by generat- ing multiple focus points, controlling intensity distribution, and preventing irradiation in neighboring region. We further extend our work to apply multiple image quality metrics with SBGA and devise a method that optimizes signals from multiple fluorescent targets at a time. By applying a simple image analysis technique, thresholding, the image is ini- tially segmented into background and object pixels. Two well known image quality metrics, entropy and mean intensity of the thresholded image are then employed as the objective functions to perform optimization. We demonstrate that combination of this method with Bessel-Gaussian beam (BG) yields higher enhancement compared to Gaussian beam. All of these methods can potentially be applied to develop noninvasive or minimally invasive optical tools for medical diagnosis or treatment.
Description
May2024
School of Engineering
School of Engineering
Full Citation
Publisher
Rensselaer Polytechnic Institute, Troy, NY