Known methods of heart function analysis, while useful, suffer from a number of limitations. These include the fact that these methods are inherently static, in that they consider an individual electrocardiogram (ECG) that can only provide a picture of heart function at a particular moment in time. Furthermore, the assessment depends upon the expertise of specialists performing a visual analysis of ECG traces, which may be subjective, and of limited sensitivity. Accordingly, a method of processing ECG signals comprises the steps of extracting data from at least one ECG signal, determining a time derivative of the data and generating a normalised index value representative of the energy in the time derivative. The index value is compared with corresponding historical data of the patient. The index may be used as an objective indication of heart function, and recorded for an individual patient over a period of time as a dynamic record of any change in heart function of the patient that may be indicative of abnormal function or heart disease.