Files in this item



application/pdfLE-THESIS-2016.pdf (4MB)
(no description provided)PDF


Title:Image processing for biological experiments on a mobile phone
Author(s):Le, Huy Minh
Advisor(s):Lumetta, Steven S.
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Image Processing
Biological Experiments
Thin layer chromatography (TLC)
Enzyme-linked immunosorbent assay (ELISA)
Abstract:Smartphone capabilities have been expanding rapidly since Apple Inc. introduced the iPhone in early 2007. Today’s smartphones offer decent computational power, good internet connectivity and high- pixel-count cameras. In this thesis, we leverage these resources in novel ways, using a smartphone CMOS camera as a sensing device and performing biological analyses immediately on the smartphone. We show two novel ways to exploit the capacities of the smartphone and discuss the challenges that we faced in the development of these applications. We first demonstrate a mobile system to identify counterfeit and substandard drug products effectively and inexpensively. Our system costs roughly 250 dollars, which is affordable in developing countries in Africa and Asia, where counterfeit and substandard drug products are flooding the markets. The system also enables analyses at the point of testing, which is particularly valuable when laboratory facilities are remote or unavailable. The system consists of a 3D-printed cradle, a cheap Android based smartphone and an UV lamp. The system is inspired by the use of the thin layer chromatography (TLC) method, which is known to be efficient in verifying the identity of drug products from unknown sources. The core analysis of a TLC plate is performed though a series of image processing algorithms on the Android smartphone. For drugs with a single active pharmaceutical ingredient (API) that absorbs ultraviolet (UV) light, the mobile phone-based detection system is able to discern 5% drug concentration differences, which is equivalent to a commercially available lab-based desktop TLC reader priced at roughly $40,000. We then demonstrate absorption-based biological assays by performing an enzyme-linked immunosorbent assay (ELISA) experiment. ELISA is an assay technique used to detect and quantify substances such as proteins, antibodies, hormones, etc. When combined with some simple optical components, the rear-facing CMOS camera in a smartphone can capture spectral data for biological samples. We developed image processing algorithms to calibrate and to analyze these spectra to match the results produced by conventional laboratory instruments. In order to enable unskilled users to perform ELISA experiment accurately, we integrated these techniques into an Android application with a simple user interface that walks users though assay steps. Finally, we generalize some of the lessons learned and challenges faced during development of the TLC and ELISA applications in order to provide a broader and more useful picture for developing smartphone bioassay applications in general.
Issue Date:2016-12-09
Rights Information:Copyright 2016 Huy Le
Date Available in IDEALS:2017-03-01
Date Deposited:2016-12

This item appears in the following Collection(s)

Item Statistics