Breast cancer evaluation by fluorescent dot detection using combined mathematical morphology and multifractal techniques

Abstract

BackgroundFluorescence in situ hybridization (FISH) is very accurate method for measuring HER2 gene copies, as a sign of potential breast cancer. This method requires small tissue samples, and has a high sensitivity to detect abnormalities from a histological section. By using multiple colors, this method allows the detection of multiple targets simultaneously. The target parts in the cells become visible as colored dots. The HER-2 probes are visible as orange stained spots under a fluorescent microscope while probes for centromere 17 (CEP-17), the chromosome on which the gene HER-2/neu is located, are visible as green spots.MethodsThe conventional analysis involves the scoring of the ratio of HER-2/neu over CEP 17 dots within each cell nucleus and then averaging the scores for a number of 60 cells. A ratio of 2.0 of HER-2/neu to CEP 17 copy number denotes amplification. Several methods have been proposed for the detection and automated evaluation (dot counting) of FISH signals. In this paper the combined method based on the mathematical morphology (MM) and inverse multifractal (IMF) analysis is suggested. Similar method was applied recently in detection of microcalcifications in digital mammograms, and was very successful.ResultsThe combined MM using top-hat and bottom-hat filters, and the IMF method was applied to FISH images from Molecular Biology Lab, Department of Pathology, Wielkoposka Cancer Center, Poznan. Initial results indicate that this method can be applied to FISH images for the evaluation of HER2/neu status.ConclusionsMathematical morphology and multifractal approach are used for colored dot detection and counting in FISH images. Initial results derived on clinical cases are promising. Note that the overlapping of colored dots, particularly red/orange dots, needs additional improvements in post-processing.

Application of multifractal analysis on microscopic images in the classification of metastatic bone disease

“…A smaller number of studies researches possibilities (for Ref. see Mukundan and Hemsley 2010) of applying multifractal parameters in the classification of medical images and thus in the medical diagnostics (Reljin et al 2002(Reljin et al , 2009(Reljin et al , 2011Bartnykas 2010). The results of these studies point to a field of great potential for further study in order to increase the reliability of cancer diagnosis and its acceleration and facilitation.…”

“…For instance, high values of Holder exponent α denote high local changes, and opposite for low α. Regarding to the MF spectrum f (α), its low values denote rare events -isolated parts in the whole structure having particular value of α, and opposite for high f (α) (Reljin 2011). The multifractal theory and the associated multifractal spectrum are useful for describing the irregularities of biomedical images.…”

“…Artificially generated fractal structures are commonly known as deterministic (or, mathematical) fractals (Mandelbrot 1982;Evertsz and Mandelbrot 1992). These structures are generated by using exact rules and they are characterized by exactly the same fractal dimension in whole scales, thus they are referred as monofractals (Reljin et al 2011).…”

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

“…While the technological and computational development continued with undiminished energy during the following years (22,35,38,41,44,46,60,61,66,(81)(82)(83)(84)(85)(86)(87)(88)(89)(90)(91)(92)(93), even more emphasis was laid on practical approaches and a broad spectrum of clinical applications appeared in the beginning of the 21st century.…”

Image thresholding techniques for localization of sub‐resolution fluorescent biomarkers

“…These signal enhancement techniques are often based on wavelets (22) and mathematical morphology (MM) (23)(24)(25)(26). In particular, the MM-based top-hat (TH) filter is widely used for removing low-frequency contributions such as background fluorescence, out-of-focus fluorescence, and cytoplasm auto-fluorescence (18,19,27,28). Aside from global thresholding, local thresholding techniques are also used for extraction of spots (29) by computing a specific threshold for each pixel based on its surroundings.…”