An automated DNA strands detection system featuring 32-bit ARM7TDMI microcontroller and VGA-CMOS digital image sensor

Loading...
Thumbnail Image
Date
2009-03
Authors
Mohd Noor, Mohd Halim
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Genetic DNA recognition is a routine experiment for detecting the origin of the species. Electrophoresis is one of the processes for such detection which has been used extensively. One popular technique based on electrophoresis is agarose gel sequencing. This method separates DNA molecules by size and eventually produces high resolution DNA strands. It is achieved by moving the negatively charged molecules through an agarose matrix towards cathode and anode if the molecules are positively charged. Manual analysis on the agarose gel image, usually is time consuming and prone to human errors. Automated agarose gel image analYSIS nullifies the problems and speed up the process. Hence the objective of this project is to develop an imaging system for automated DNA strands detection. This project can be divided into two stages. The first stage is the development of an imaging system that is able to capture agarose gel electrophoresis images. An electronic circuit is designed to interface and control two main components of the imaging system which are CMOS image sensor and microcontroller. The design is focused on developing a memory system that can temporarily store the image data before it is used for image processing. The imaging system was tested on a real electrophoresis application. The system is successfully able to capture agarose gel electrophoresis image of size 320 x 240 of 10-bit length. It was observed that the images were heavily corrupted with noise inherently present in the imaging system. The second stage is the development of image processing algorithms and pattern recognition software. The DNA detection algorithm adopts image subtraction technique to establish DNA recognition. For simplicity the algorithms are implemented in MATLABĀ® environment. The image comparison algorithm is based on image subtraction technique. The reference image is subtracted with the sample image and the maximum intensity value and average of intensity value of the resultant image are determined. The values are compared against the threshold values to establish recognition. Twenty samples of five types of DNA have been used to evaluate the accuracy and fidelity of the proposed procedures and methods. It was found that the system is able to classify DNA strands with 98.16% of average accuracy rate.
Description
Keywords
Genetic DNA recognition is a routine experiment , for detecting the origin of the species
Citation