A New Method for Calculating Lane Average Width on the Pulsed Field Gel Electrophoresis (PFGE) Images for Lane Detection and Extraction Problem

We aimed to a new method to calculate the lane average width on the pulsed field gel electrophoresis (PFGE) images for lane detection and extraction problem. Although some studies are reported for lane detection based on vertical projection profile, they are not automatic with low error. Average width of lane is the most important parameter that is required for automatic image processing of PFGE images. This research with the aim of using the power spectrum density to calculate the lane average width was carried out. First, based on the power spectral density, PFGE images were processed. The proposed algorithm was trained using 10 PFGE images and then evaluated for 20 PFGE images which totally consisted of 300 lanes. These images were developed using Bio-Rad model of PFGE in Microbiology Laboratory of Kermanshah University of Medical Sciences, Iran. The power spectrum density procedure in contrast to intersection of the horizontal lane yields decreased 99.61% of calculation error for lane detection. Considering the lane average width is used in several stages for lane detection and extraction procedure, it can be concluded that the power spectrum density improves the process of lane extraction significantly.