Effective Automatic Recognition of Cultured Cells in Bright Field Images Using Fisher’s Linear Discriminant Preprocessing
Recent progress in the development of methods for molecular genetic analysis (e.g. RT-PCR, microarrays) has brought sensitivities to the level where single cells can be analyzed [1]. To carry out single-cell-level assays on significant numbers of cells, high throughput robotic systems are needed. These systems require identification of cultured cells (often in bright field images) for micromanipulation and subsequent molecular analysis. Given the variability of cell size and morphology, the presence of “trash,” as well as variations in microscope parameters, such as focus and illumination, identification of cultured cells in bright field images is a difficult task that, traditionally, is done by an experienced human observer. However, the use of human observers represents a severe impediment to the development of high throughput robotic systems. Therefore, there is a major need for algorithms that permit automatic recognition of cells in bright field images.