Abstract:
The principal objective of the present work is to understand the elasticity imaging complexity and the probability of noise to measure Young‟s modulus based on number, size and location of tumors. This study is to make the modulus reconstruction process robust. For numerical computation, a model of breast tissue with tumor inclusion is created using a finite element method based modeling software called COMSOL Multiphysics. Fat and skin are placed to make the model more realistic. This is a 2-dimensional model. 3cm by 3.6cm area of soft tissue is modeled first. Tumor of fixed radius is included then. The Young‟s modulus values for fat, skin, soft tissue and hard tumors are set according to literature. More models are designed for different number, size and location of tumors. These models are used for error analysis of elasticity imaging technique. For elasticity imaging it is necessary to press a probe on the top surface which is implemented by fixed displacement in our model. Displacement is set according to literatures. Values of stress and strain of 27462 points of the model are exported from Comsol Multiphysics to Matlab for calculation of Young‟s Modulus. A matrix of 361 by 301 is made with interpolated values in Matlab for further calculation. Percentage error is calculated for seven sections of the model (one inside tumor, six outside tumor). It is found that error is less near to the upper surface, error increases with increment in depth. Error is less inside tumor comparing to outside tumor. Maximum error is 188.48 at lower side and minimum value is 13.3 at Inside tumor. For various tumor sizes, error increases with increase of tumor size. But inside tumor error decreases with increase of tumor size. For double tumors it is found that error is reduced near to the small tumor. when the small tumor is at top lower side error is 77.62% but when the smaller tumor is at bottom the error is less and it decreases to a value of 58.68%. There is not much effect for variation of displacement field rather than 0.03 cm for tumor at middle of the tissue model. So, the numerical simulation can be viably used to develop a mathematical error pattern which can help understand the elasticity imaging complexity and the probability of noise to measure Young‟s modulus and hence can improve efficiency of modulus reconstruction process.
