This paper describes the real-time implementation of a speech fundamental period estimation algorithm known as MLP-TX. The algorithm uses a multi-layer perceptron (MLP) classifier to determine the location of the points in time of vocal fold closure in a noisy speech signal. The real-time implementation was carried out in assembly language using a TMS320C25 signal processor. The system will be used in the next generation of SiVo hearing aids developed by the EPI group at UCL. These aids provide voice frequency information as a guide to Lipreading and voice control for the profoundly deaf.