AN AUTOMATIC ALGORITHM FOR IDENTIFICATION AND STRAIGHTENING IMAGES OF CURVED HUMAN CHROMOSOMES

Abstract

Karyotyping is a standard method for presenting the complete set of the pictures of human chromosomes in a table-like format. It is usually used by a cytogenetic expert to predict the common genetic abnormalities. Producing a Karyotype from microscopic images of human chromosomes is a tedious and time-consuming task, so an automatic Karyotyping system would help the cytogenetic expert in his/her routine work. Automatic Karyotyping algorithms usually suffer the non-rigid nature of the chromosomes, which makes them to have unpredictable shapes and sizes in the images. One such problem that usually needs the operator's interaction is the existence of curved chromosomes within the images. In this paper, an effective algorithm for identification and straightening of curved human chromosomes is presented. This will extend the domain of application of the most of the previously reported algorithms to the curved chromosomes. The proposed algorithm is applied to single chromosomes that are initially modified by means of a Median filter. The medial axis (MA) of the filtered image is then extracted using a thinning procedure, which is carried out on the binary version of the image. By comparing the Euclidean distance of the endpoints and the length of the MA, a curved chromosome is identified. For chromosome straightening, the initially extracted medial axis is then modified by extending it in both ends considering the slope of the MA. Next, the original input image is intensity sampled over many closely located perpendicular lines to the MA along the chromosome which are then mapped into a matrix (as rows) producing a vertically oriented straight chromosome. For evaluation, the algorithm is applied to 54 selected highly curved chromosomes obtained at the pro-metaphase stage, which were provided by the Cytogenetic Laboratory of Cancer Institute, Imam Hospital, Tehran, Iran. The density profile and the centromeric index of the chromosomes which are among the most important and commonly used features for chromosome identification are calculated by the expert both before and after the straightening procedure. The mean squared error and the variance of the difference between the two are then obtained and compared. The results show a good agreement between the two, hence the effectiveness of the proposed method. The proposed algorithm therefore extends the domain of application of the previously reported algorithms to the highly curved chromosomes.