Detection and characterization of antinuclear antibody using fluorescence image processing

The production of autoantibodies in the human cells (stomach, liver) is detected by indirect immunofluorescence assay and each disease condition is indicated by its respective fluorescence pattern during fluorescence microscopy. There are many different patterns which have close resemblance and a major hindrance in the disease diagnosis is the distinction among fluorescence patterns that have close resemblances with minor differences. Disease conditions vary with intensity, type and localization of fluorescence in each specific region. An algorithm using MATLAB has been developed which will facilitate auto-detection of pattern with respect to its corresponding control pattern thus assuring comparison with correct match and minimizing the chances of disease misdiagnosis. Future application of this methodology is in karyotyping and detection of antibodies through fluorescence imaging employed in diagnosis of variety of diseases (including cancer).