IMAGE-ANALYSIS FOR COMPARATIVE GENOMIC HYBRIDIZATION BASED ON A KARYOTYPING PROGRAM FOR WINDOWS

OBJECTIVE: To describe the image processing and analysis techniques we developed for the quantitative analysis of comparative genomic hybrid ization (CGH). STUDY DESIGN: A system for CGH cytometry is based on a semiautomated karyotyping program using the Windows graphic user inter face. RESULTS: After alignment and normalization of the fluorescence i mages of the test genome (FITC) and reference genome (TRITC), the chro mosomes are segmented and arranged as a CGH karyogram. The CGH karyogr ams from different metaphases of one tumor sample are represented as a CGH sum karyogram. Mean DAPI, FITC, TRITC and RATIO images, as well a s ratio profiles with the 95% confidence interval, can be displayed. C ONCLUSION: Representation of CGH results in the form of pseudocolored mean ratio chromosomes enhances the visibility of the method. The rati o profile plus confidence interval facilitates the identification of p reparation artifacts and the classification of chromosomal imbalances. The sum karyograms of several tumors can be combined into a CGH super karyogram of a tumor subgroup; that helps identify recurrent DNA chang es in a tumor and might lead to genetic tumor classification based on CGH.