Measuring mechanical strain is crucial in several engineering applications. Ultrasonic guided waves can be used to measure strain since the propagation speed is a function of the stress-state. Strain monitoring is usually performed by observing the time-of-flight shift. However, in complex media, interpretation of signals can be difficult, given the effects of dispersion, for example, causing several wave modes propagating simultaneously. In these cases, the time-reversal or the cross-correlation techniques can be used to monitor the strain level by observing the peak of the focused signal which changes proportionally to the strain level. However, not all components of the spectrum contribute with the same sensitivity to strain changes. In this capstone project, this process is modified by using a new signal as reference which is synthesized through a filtering procedure based on the phase of the Fourier, the short-time Fourier and the Wavelet transforms spectra, relying on prior knowledge of the system impulse response at some non-null strain level. The technique was evaluated numerically in signals obtained from an aluminum plate, effectively producing more strain sensitive signals. The sensitivity increase depends on a phase threshold parameter used for the filtering process. The technique was assessed based on the sensitivity gain, the loss of energy concentration capability and the value of the foreknown strain. Signals synthesized through the time frequency representations, either the short-time Fourier transform or the Wavelet transform, provided a better tradeoff between sensitivity gain and loss of energy concentration.