Many physiologic indicators can help a physiologist to interpret and predict how athletes will react to a certain load. These can be divided into two major categories: (i) invasive methods, and (ii) non-invasive methods. The former can only be assessed with blood parameters, being impractical in a dynamic environment at the current state of the technology. As this book revolves around the topic of pattern recognition in sports during training and competition, the focus will be placed on the latter. In non-invasive physiological methods, one can assess many variables that can help to better understand how the athlete is reacting to the loads without blood parameters and, at the same time, maintaining the ecological circumstances of performance in obtaining the data by employing technologies that do not perturb the athlete. This section includes the description of basic physiological features, such as heart rate and muscle activity (electromyography). These are examples of physiological data generated by the human body from which it is possible to assess a person’s stress, which may be perceived not always consciously nor through non-physiological data (Jerritta, Murugappan, Nagarajan, & Wan, 2011). All this is possible due to the ability to draw more representative features from these signals, such as the mean heart rate variability, muscle load, and other related metrics obtained through signal processing, such as electromyography root mean square and electromyography Fourier transform.
Heart rate (HR) reflects the number of systoles that occurs per minute, caused by the activity of the parasympathetic and sympathetic nervous system in the sinus node. These systems work in search of balance, since the sympathetic system is responsible for the body’s responses to certain impulses, making the heart rhythm to meet the necessary requirements. The parasympathetic system, however, is responsible for the heart response when at rest (HR between 60 and 100 beats per minute) and also the one that assumes the relaxation function by slowing the HR when facing an increased activity.
The resting heart rhythm (HRmt) represents the basal heart rhythm and should, therefore, be measured after awakening and preferably in an isolated space where the person suffers as little disturbance as possible. In athletes, this tends to decrease with aerobic training, so it is of high importance to know this metric for each athlete in order to be able to evaluate his/her performance or to make the screening of any health problem. The maximum heart rhythm (HRnulx) is reached when the athlete is under a high physical load, forcing the body to request as much oxygen as possible, causing the number of systoles to reach a maximum. Through these two concepts, it is possible to obtain the heart rate reserve (HRr) for player i as follows:
Given the current heart rate values of a player i, HR, [f], which can be extracted with state-of-the-art heart rate monitors (see Chapter 3), and considering that HR,MXi and HRrcsli are known constant values, the heart rate values can be normalized for each time step t by applying the following equation:
Besides heart rate, the heart rate variability (HRV) is often adopted to monitor training loads. The analysis of HVR can be made by computing a time series of RR intervals, i.e., the intervals between successive heartbeats, or R waves. If updated at every timestamp t for each player i, the time series RR can be expressed as a time-varying variable RR, [г]. This can, once again, be tracked with wearable technologies, such as heart rate monitor chest bands. With RR, [/] as an available feature, one can calculate multiple additional physiological metrics. These metrics, however, since they rely on average measurements, shall only be updated at every integer number of steps t (e.g., every second). The mean heart rate variability (mHRV, [f]) is the most common metric, being calculated as the simple moving average of RR, [f]:
The standard deviation of all NN intervals (normal RR intervals), known as SDNNj [/], indicates the global HRV and can be calculated as follows:
The root mean square of the successive differences in RR intervals (RMSSDj [(]) allows estimating variations in heart rate in short-term RR recordings, being calculated as follows: