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

Abstract

The paper considers the method, based on multifractal (MF) analysis, for classifying the shape of tissue cells from microscopis images, identifying the primary cancer in cases of metastasis bone disease. Diagnosis of primary cancer is of great importance, because further treatment depends on how successful and accurate that diagnosis is. This method can be applied as an additional and objective tool in primary cancer diagnosis, as well as in decreasing of the subjective factor and error probability. The method is tested over a large number (1050) of clinical cases from the Institute of Pathology, University of Belgrade. The results of computer-aided analysis of images have been presented and discussed.

Music Performers Classification by Using Multifractal Features: A Case Study

“…Since the MF analysis has been proven as a powerful method for finding subtle details within signals (Gavrovska et al, 2013) and for characterizing and distinguishing different signals (Reljin et al, 2008;Vasiljevic et al, 2012) we investigated the MF approach for resolving given problem. As examples we will observe the same 14 songs, listed in Table 1, performed by famous Swedish group ABBA, producing mega hits from 1974 to 1982 (Sheridan, 2012), and by actors in the movie Mamma Mia (released in 2008).…”

“…For instance, MF was applied to find clicks in heart sounds characterizing pathological syndrome so-called the MVP (mitral valve prolapse, or click-murmure syndrome), as reported in (Gavrovska et al, 2013). Moreover, it was shown that some MF parameters can be used as characteristic features for discriminating malignant from benign cases from appropriate medical signals, as shown in (Reljin et al, 2008), or for identifying the primary cancer from biopsy images of bone metastases (Vasiljevic et al, 2012).…”

Application of multifractal analysis to segmentation of water bodies in optical satellite images

“…Nilsson [21] studies the possibility of using this methodology to distinguish healthy tissue from a potentially tumoral one based on the differences in the width of the corresponding spectra. Vasiljevic et al [22] use particular parameters of multifractal analysis, like some generalized dimensions or the range of Hölder exponents, for classifying the primary cancer in cases of metastasis bone disease.…”

Medical Imaging Mutifractal Analysis in Prediction of Efficiency of Cancer Therapy

“…In parallel to standard molecular approaches, digital pathology emerged as a structure analysis tool to aid in detection, diagnosis (Cross 1997;Vasiljevic, Reljin et al 2012), risk assessment (Vasilescu, Giza et al 2012), chemotherapy efficacy assessment (Li, Hu et al 2014) and therapy prediction for cancer (Laurinavicius, Plancoulaine et al 2014). It is based on computational analysis of medical images by use of texture or fractal algorithms.…”