Automatic identification of chromosomal abnormalities in metaphase karyotype using paired images in human chromosomes

Identification of chromosomal abnormalities is necessary part of diagnosis and treatment of genetic disorders and some cancers. Cytogenetic is a preferred tool in the diagnosis of genetic diseases. In cytogenetic, karyotype which is a systematic representation of human chromosomes is obtained through imaging a nucleus of cell using optical microscopes. And undergoes an analysis in which chromosomes are sorted based on morphological features. At presents this analysis is carried by lab technician through eye technique which is time-consuming and prone to error. This research aims to find automatic identification of chromosomal abnormalities in human karyotypes. The main step in main automation of this method is to define some morphological features for each chromosome. In this method, we can find the presence or absence of abnormalities or chromosomal deviations by determining the position of centromere automatically with length, the ratio of short arm to long arm of each chromosome and its comparison with the range of standard deviation in normal chromosome. This method was tested on 40 karyotype images of patients and all types of abnormalities were detected in human karyotypes, except for cases of displaced abnormality between two pieces of chromosome with similar size in two similar arms (P or q).