This paper is focused on presenting Multi-level decomposition analysis for foetal Electrocardiogram (FECG) signals, in order to improve non-inasive methods of treatment during the prenatal stage of pregnancy. Multi-level theory is presented herewith different bases of wavelets such as Daubechies and Haar by reconstructing foetal heart rate (FHR) with denoised wavelet coefficients. 4th Daubechies was used for denoising and Haar basis for foetal heart beats detection, due to the differences between maternal and foetal heartbeats and the similarity of Daubechies basis wavelets to the complex QRS of a heartbeat. It was possible to reconstruct the signal and furthermore analyze only the most important parameters of the waveform. This procedure did an effort to improve FECG analysis using a low computational mathematical tool, nevertheless it is still being improved.