Soft tissue characterization using genetic algorithm

Soft tissue properties are important to many modern applications of technology to medicine, such as robotic surgery, soft tissue modeling, and surgical simulation with force feedback. However, realistic acquisition of soft tissue properties is extremely challenging, not only because of the nonlinearity, anisotropy, nonhomogeneity, rate dependence, and time dependence of soft tissues but also due to the layered and nonhomogeneous structures of soft tissues (Samur et al. 2007; Kim et al. 2008; Zhong et al. 2010, 2012). It is understood that soft tissue properties may dynamically change during the surgical process according to different patients, different organs, different functional regions and layers crossed by the surgical tools, and different physiological conditions. Therefore, it requires that mechanical properties of soft tissues be acquired and studied through a real-time intraoperative measurement process. Currently, soft tissue properties are commonly acquired by conducting standard mechanical testing such as tensile and indentation tests on material samples extracted from organs under well-defined loading and boundary conditions (Yamada and Evans 1970; Fung 1981; Woo et al. 1993; Ahn and Kim 2009; Ahn et al. 2012; Koo et al. 2012). Since the thickness and transversal area of organs are not easy to measure, strain and stress values cannot be obtained directly from the measured displacement and force data (Samur 2007; Chawla et al. 2009). As such experimental conditions are impossible to achieve, the parameter estimation methods may not be representative when they are determined based on in vitro experiments (Kohandel et al. 2008).