Analysis of PCR-RFLP gel electrophoresis images for accurate and automated HPV typing

The identification of the types of the human papilomavirus (HPV) that have infected a woman provides valuable information as regards to her risk for developing cervical cancer. HPV typing is often performed by means of manually analyzing PCR-RFLP gel electrophoresis images. However, the typing procedure that is currently employed suffers from unsatisfactory accuracy and high time consumption. In order to treat these problems we propose a novel approach to HPV typing that automates the analysis of the electrophoretic images and concurrently improves the accuracy of the typing decision. The proposed methodology contributes both to the extraction of information from the images through a novel modeling approach and also to the process of making a typing decision based on the above information by the introduction of an original HPV typing algorithm. The efficiency of our approach is demonstrated with the help of a complex worked example that involves multiple HPV infections.