Quantification of effects of cancer on elastic properties of breast tissue by Atomic Force Microscopy

Different behaviors of cells such as growth, differentiation and apoptosis widely differ in case of diseases. The mechanical properties of cells and tissues can be used as a clue for diagnosis of pathological conditions. Here, we implemented Atomic Force Microscopy to evaluate the extent of alteration in mechanical stiffness of tissue layers from patients affected by breast cancer and investigated how data can be categorized based on pathological observations. To avoid predefined categories, Fuzzy-logic algorithm as a novel method was used to divide and categorize the derived Young's modulus coefficients (E). Such algorithm divides data among groups in such way that data of each group are mostly similar while dissimilar with other groups. The algorithm was run for different number of categories. Results showed that three (followed by two with small difference) groups categorized data best. Three categories were defined as (E < 3000 Pa, 3000<E <7000 Pa and E > 7000 Pa) among which data were allocated. The first cluster was assumed as the cellular region while the last cluster was referred to the fibrous parts of the tissue. The intermediate region was due to other non-cellular parts. Results indicated 50% decline of average Young's modulus of cellular region of cancerous tissues compared to healthy tissues. The average Young's modulus of non-cellular area of normal tissues was slightly lower than that of cancerous tissues, although the difference was not statistically different. Through clustering, the measured Young's moduli of different locations of cancerous tissues, a quantified approach was developed to analyze changes in elastic modulus of a spectrum of components of breast tissue which can be applied in diagnostic mechanisms of cancer development, since in cancer progression the softening cell body facilitates the migration of cancerous cells through the original tumor and endothelial junctions. (C) 2015 Elsevier Ltd. All rights reserved.