It is presented the partials results of an early uterine cancer detection based on electrical impedance spectrum. It was developed a new algorithm to the inversion of the Cole-Cole model using only the real part of the electrical impedance spectrum. The new algorithm is based in a hybrid inversion model, a direct inversion and a genetic algorithm. The basic procedure is described by the following sequence. First, the imaginary part of the electrical impedance spectrum is precalculated using a direct inversion of Debye model. Second, the Xiang direct inversion is used to calculate a first approximation of the Cole-Cole parameters. Finally, it is use of a genetic algorithm to optimize the Cole-Cole parameters doing minimum the square error between the data and the model.