Heart auscultation (the interpretation by a physician of heart sounds) is a fundamental component in cardiac
diagnosis. It is, however, a difficult skill to acquire. In this work, we present a study for a system intended to aid in
heart sound classification based on a FFT of the sounds and a neural network. This work contain three step. The
information acquire is the first step which included recording the heart and can record by small recode instrument.
The sound from the patient by the sonoketle phone , where the sound heart can be heard and can record by small
recoed instrument _ the second step is analysis step ,in this step we analysis the sound wave file to get (11)
parameter which represent the input to to the third step (classification step) .
ln classification step we can recognize the class which the sound wave file belong to it. heart sounds of 21
subjects divided in to groups normal (6) and heart valve diseases (l4)analysis and take times and frequencies as 11
parameters that interred to input of network . The accurate result was obtained.