Measurement of chaos and quantum decoherence in H&E stained images of oral epithelium for early detection Oral Submucous Fibrosis

Abstract

Oral cancer is accountable for 0.3 million deaths annually, and in India a significant factor which adds to this burden is the delayed detection of precancerous states like Oral Submucous Fibrosis (OSF). Histopathological abnormal changes seen in the oral epithelial region is known as dysplasia which is accompanied by abnormal variation in nuclear and cellular size and shape, atypical mitotic figures, and augmented nuclear-cytoplasmic ratio. To diagnose dysplastic changes in histopathological images, computer aided image processing is being used extensively. The present study measured Fractal dimensional, power spectral density and Renyi entropy values of the epithelial nuclei for the different study groups i.e. Normal, OSF (without dysplasia) and OSF (with dysplasia) to understand chaos whereas detection of the number of mitotic figures was done to understand quantum coherence. The analysis were done ImageJ and MATLAB respectively. Fractal dimensional analysis demonstrated significant increase (p<0.0001) in diseased conditions. Renyi entropy analysis, found an increase (p<0.00001) in both the precancerous conditions. The power spectrum decay rate shows increase in chaos from normal to precancerous stage with dysplasia (p=<0.00001) and also between the precancerous conditions (with and without dysplasia (p=0.00028). Mitotic figures were successfully detected using combination of deep blue ratio and contour detection. An enhancement in number of mitotic figures from standard to precancerous condition was observed and hence loss of quantum coherence was established in diseased conditions.