In order to fix the problem of ill-posedness

some a priori information should be incorporated in the process of optical tomography reconstruction.In this paper

a Gibbs distribution with binary line process is introduced as a prior image model

which can result in a global smoothness with sharp edges.Because of the coexistence of the binary and continuous variables in the objective function

traditional optimization algorithms are not valid.Therefore

a coupled gradient neural network is proposed.In the process of optimization

the gradient computation of the energy function with respect to optical parameters is critical

for which

an algorithm based on the gradient tree is put forward.The reconstruction images corresponding to both the absorption and scattering coefficients proved that the proposed algorithm can be implemented effectively with high quality results by the introduction of the binary line process.