Another type of global attribute is a conventional intrinsically non-temporal feature. For example, consider diagnosing medical patients by observing the electrical patterns and behaviour of their heart - in other words, an electrocardiograph (ECG) - as discussed in Section 6.3.3. Obviously there are important temporal characteristics of the ECG; for example, the rates of change of the electrical signal, the order of various sub-components of the heart's beats and so on. Aggregate temporal attributes as outlined before are also useful. But there are also other attributes which are important to classification and diagnosis, such as the patient's age, height, weight and gender. These are not directly temporal measures, but they do have impact on the diagnosis. One of the advantages of the TClass approach is that it allows the integration of conventional features, aggregate global features and temporal features while other systems (such as dynamic time warping and hidden Markov models) do not. The practicalities of such integration are the topic of the next section.