An inverse problem formulation for parameter estimation of a reaction-diffusion model of low grade gliomas (Q907131)

The authors present a numerical scheme for solving a parameter estimation problem for a model of low-grade glioma growth. The goal is to estimate the spatial distribution of tumor concentration, as well as the magnitude of anisotropic tumor diffusion. Mathematically, the problem consists in reconstructing the scalar coefficient \(k_f\) and the initial state \(c_0(x)=c(x,0)\) in the boundary value problem \(\partial c/\partial t-Dc-R(c)=0\) in \(u=B\times (0,1)\), \(\partial c/\partial n=0\) on \(\Gamma\times (0,1)\). Here, \(B\) is a spatial domain, \(\Gamma=\partial B\), \(Dc=\nabla (k(x)\nabla c)\) with \(k(x)=k_0(x)I+k_f T(x)\), \(R\) is a nonlinear operator. As observation data, the authors choose indicator functions of the sets \(\{ x: c(x,t)\geq c_d \}\) for \(t=0, 1\) with a given threshold \(c_d>0\). The Tikhonov regularization method for solving the inverse problem is applied. Numerical results with a synthetic data for different noise levels and detection thresholds are presented.