Fluorescent dot counting in interphase cell nuclei

Abstract

Fluorescence in situ hybridization allows the enumeration of chromosomal abnormalities in interphase cell nuclei. This process is called dot counting. To estimate the distribution of chromosomes per cell, a large number of cells have to be analysed, particularly when the frequency of aberrant cells is low. Automation of dot counting is desirable because manual counting is tedious, fatiguing, and time consuming. We have developed a completely automated fluorescence microscope system that counts fluorescent hybridization dots for one probe in interphase cell nuclei. This system works with two fluorescent dyes-one for the DNA hybridization dots and one for the cell nucleus. A fully automated scanning procedure has been used for the image acquisition. After an image is acquired it has to be analysed in order to find the nuclei and to detect the dots. This article focuses upon the dot detection procedure. Three different algorithms are presented. The problems of 'overlapping' dots and split dots are discussed. The automated dot counter has been tested on a number of normal specimens where DAPI was used for the nucleus counter stain and a centromeric probe was used to mark the chromosome 12. The slides contained lymphocytes from cultured blood. The performance of the different algorithms has been evaluated and compared with manually obtained results. The automated counting results approximate the results of manual counting.

Algorithms for Cytoplasm Segmentation of Fluorescence Labelled Cells

“…A method inspired by the watershed algorithm [13,20] was used for initial separation of clusters of cells and segmentation of the image into cells and background in one step. A 2D grey-level image can be thought of as a topographic relief where the cells with high intensity are peaks separated by lower intensity valleys.…”

State‐of‐the‐art FISHing: Automated analysis of cytogenetic aberrations in interphase nuclei

“…Since the size, shape, and intensity of signals can differ significantly even within the same nucleus (64), separating them from artifacts and aspecific background noises is not always possible. Further sources of difficulties are the insufficient hybridization, presence of split signals, and random colocalization of signals with the same color (17,65). When investigating nuclei harboring highly amplified genes, nondot-like signals (clusters) generate problems with standard spot counting methods; these are often presented in tissue samples (e.g., amplification of HER2 and N-myc genes) making analysis even more difficult.…”

“…Although, even nowadays, CISH still is a technique in its own right, the main attention quickly returned to fluorescence applications. In 1997 and 1998, two milestone papers were published about automated i-FISH analysis by Netten et al (65,75). These papers still often serve as reference points for system development because they provide comprehensive and detailed descriptions about each step of dot counting; furthermore, crucial problematic issues raised from sample quality, e.g., low cell density, inhomogeneous cell distribution, split and overlapping signals, are also addressed.…”

Accurate and Fast Off and Online Fuzzy ARTMAP-Based Image Classification With Application to Genetic Abnormality Diagnosis

“…The features include Area (a size measure) and Eccentricity (a shape measure), which were previously suggested [19] to describe FISH signals. In addition, we measured a number of spectral features.…”