Hello and welcome back. We're now going to take a deeper dive into the relationship that internal and external measures have with each other with a focus on training. By way of review of our previous topics, we have talked about how external measures are mainly derived from accelerometer or GPS data to quantify the movement experienced by the athlete or individual wearing the device. Separately, internal sensors are able to provide us a view of several types of physiologic measures that are occurring within the body, such as heart rate, heart rate variability, and skeletal muscle oxygenation. What is the benefit of combining the internal and external measures? Well, access to both the external and internal sensor measures allows for a deeper understanding of the training effort. It's relative intensity, absolute intensity, and its relation to previous training efforts. Capturing the external measures, for example, player load or player load per minute, or GPS velocity or distance provides us information about what the athlete did or is currently doing while comparing it with the internal measures which describe their effort in getting it done. I feel that heart rate is still one of the best global measures of relative intensity, let's consider a different internal measure example. If we were to measure skeletal muscle oxygenation of an athlete from a muscle that is critical to their activity, for example, measuring the vastus lateralis of a runner on a treadmill, then we could begin to provide insights about the muscle specific speed thresholds for that athlete, which would help determine aerobic and anaerobic thresholds. This particular example is very specific and it seems more like something used during an exercise test. Let's first talk more about common combinations of internal and external measures. These days, a single wearable device can measure both internal and external measures now that heart rate can be measured at the wrist. A smartwatch or even a smart band will have the capacity to measure heart rate and GPS velocity. In fact, wearables can allow for the integration of many internal and external measures to be collected continuously. Importantly, internal and external measures are often included in algorithms for device-specific metrics. For example, a company named, First Beat, created a patented method for estimating fitness level using GPS velocity data and PPG heart rate data. This picture is from part of their white paper about predicting aerobic fitness from the accumulation of what we might call regular training. In this set of examples of good and bad data, we see a person that is out for a run with a device that can capture both running speed and heart rate. In the first segment of this figure, the individual is running at a low speed and they have a relatively low heart rate. We predict a linear relationship between running speed and heart rate and so a big part of the technology advancement here is determining which segments of the training are reliable and should be used in the algorithm and which are unreliable and should be excluded. What we see in the next portion of the figure is that the runner is stopped at a traffic light. So there is no running speed and a declining heart rate. This segment is not part of what would be used for determining the athlete's aerobic fitness. In the next segment of the run, however, the runner is up to a steady pace and heart rate has gone up. This is another reliable segment. Next, the runner enters a bridge and has an unreliable GPS signal. This is another segment that is excluded from the algorithm. Next, some variable pace running is performed, and this provides additional information about the speed and heart rate of the runner. A final example of unreliable data is shown here, where there is an unreliable heart rate signal resulting from poor connection between the device and the where and a final piece of helpful to data is the incorporation of fast running by the where providing perspective on the upper range of the athlete's heart rate and running speed. So an important part of the advancement here is determining whether the heart rate and the GPS data are reliable and then using the velocity in her rate data to predict aerobic fitness. The expectation is that a highly fit athletes running fast and maintaining lower heart rates compared to less fit athletes. Although the first beat algorithm has been an exciting development, especially for endurance athletes, most team sports have a more stop-and-go sprint and coasts type cadence during competitions. However, with the continuous measure, a player load and possibly GPS data, along with heart rate measures, another very relevant indicator of aerobic fitness and readiness for competition is heart rate recovery during competition. For example, in many team sports, there are all out moments where the athletes are giving all the effort that they can, followed by an opportunity for the athletes to partially recover. For example, during free throws in basketball or between plays in American football. We can use large data sets, for example, from Catapult Sports to evaluate the ability for individuals to recover from these low active or inactive periods. That is it for this brief video. Hopefully, this has been a helpful introduction of the value of continuous and integrated measures from both internal and external sensors to get the most from your wearable devices and the athletes that wear them.
