Implementing procedures for improving proton precession frequency estimation

Introducing additional signal processing procedures, a significant improvement in accuracy has been achieved when applying the least-squares-error curve-fit algorithm previously described for frequency measurements on noisy sinusoidal signals with an exponential decay in amplitude. These techniques are especially powerful when used with the free-proton precession signal, which has the above mentioned characteristics. These procedures include weighing the zero-crossing measurements by appropriate factors in order to compensate for the exponential signal decay. Additionally, a hardware demodulation process is performed on the signal prior to detection, in order to enhance the quantification accuracy and further improve the accuracy