The quality of the solutions given by the optimized Tikhonov Regularization Technique (TRT) is evaluated in comparison with the Algebraic Reconstruction Techniques (ART), for providing inverse solutions for Brain Electrical Tomography imaging. According to the results, ART algorithms for noisy data perform similarly to TRT, but require prior knowledge of the noise affecting the data. This is not a prerequisite for the TRT method. The performance of three methods used for selecting the regularization parameter t in the TRT is evaluated via computer simulations. It is shown that the use of the L-curve criterion is more reliable than the use of the Minimum-Product and the empirical Composite REsidual and Smoothing Operator (CRESO) criteria.