Concussion traumatic brain injuries and other serious athletic health issues have taken center stage in recent years due to both new research and injuries sustained by professional athletes. This new research has shown that not only is a single concussion a potential health issue but also that concussion related damage is cumulative.
There are a lot of fitness trackers on the market that athletes are using but these trackers pose several issues. Firstly most of these trackers are for fitness only and therefore are not looking at health related issues but rather how an athlete is performing over time and how they can better perform. Another issue is the individuality of such fitness trackers, these trackers are made for an individual to track their progress and therefore multiple units cannot easily be viewed by one person such as a team coach. For the trackers that are viewable by multiple people, there is often a time delay in reading the acquired data making the data un-useful in real time athlete monitoring. Lastly most if not all fitness trackers have no way of looking at things such as concussion severity, cumulative concussion damage or imminent heart attack risk.
Reason for application
My reason for applying to this challenge is largely due to my interest in sensor related technologies and the ability to integrate different sensors data to be able to help people in everyday life. With the number of emergency room visits due to sports related injuries increasing each year there needs to away to better assess the degree of these injuries are and what their long term effects are.
Going beyond the sports fields, these same technologies can be useful in older patients who may suffer similar injuries in their daily lives as an athlete on involved in a contact sport may. Therefore with the percentage of the population above 65 increasing to the point that they they will soon become the dominant demographic, we will need away to be able to care for them without the 24 hour care of a care facility or live in help.
Since it is often easier to start with a willing customer as well as a large customer base, starting in the sports arena is an obvious beginning when looking at monitoring the vital signs as well as extent of injury in a person.
A sports helmet that monitors an athlete's body temperature, heart rate and rhythm as well as head trauma would be implemented to read and transmit these measurements back to a command post. There would be the possibility of adding an oxygen sensor by the athlete's mouth or nose to monitor their O2 intake.
The health related issues I would like to tackle with this setup is the immediate health risks an athlete may face while involved in the sport. These health risks would include heat stroke, sudden heart attack and seriousness of individual concussion as well as long term cumulative concussion force sustained by the athlete.
These risks will each be monitored by an individual or group of sensors. An outline of how each symptom will be monitored as well as how it will be diagnosed will follow.
By placing a temperature sensor on or close to the temple the body's internal temperature can be monitored. This is true due to the location of the temporal artery that is located below the temple. As the temporal artery is a major artery the temperature of the blood it carries is very highly representative of the body’s internal temperature.
With the use of an accurate temperature sensor (ADT7320FBZ ) it would be possible to determine if the athletes internal temperature is reaching a dangerous level and should be removed from the sport to cool down or if other urgent actions should be taken to reduce any health risks to the athlete.
Using the ear or potentially other parts of the head, it may be possible to monitor the rhythm and electrical signals of the heart. These signals can be checked against a baseline rhythm to determine if any abnormalities are present. These abnormalities may indicate an impending heart attack, over exertion or other potential medical risks. If the system is used constantly by the same athlete it would be possible to have the system learn the individual's unique rhythm over time and then indicate when the rhythm is no longer in the normal rage. The same could be applied to the heart rate, if an athlete’s heart rate increases either at a dangerous rate or increases above a safe threshold again this could be reported to someone monitoring the athlete.
The athlete's heart signals could be monitored using an ECG front end chip (AD8232 HRM FRONT END). The signals could then be passed to a uC to do signal analysis and potently signal matching to determine if the rhythm is no longer normal. The signals could also be passed through some form of learning algorithms to help predict abnormalities. In the case of any such abnormalities being detected it would then be a simple matter to alert someone for urgent medical help.
Individual Concussion Severity
As the head is the limb most at risk due to blunt force trauma, this would be the center of the health monitoring system. With the use of two or more 3-axis accelerometers it would be possible to reasonably accurately measure the force sustained by an impact to the head. This would be an important and extremely useful bit of information in determining if an athlete should continue in the sport or be pulled to prevent further injury. In the event of a severe injury were an athlete would not be able to continue playing the information would help doctors to better understand the nature and extent of the injury sustained.
The use of two or more accelerometers (ADXL375, ADXL377, ADXL362) would be used to determine the extent of the force as well as what type of force was sustained in the head trauma. Further to this it may be possible and even useful to incorporate a gyroscope to help determine if there was a high rate of rotational force that may have caused a neck injury and to what extent this injury has occurred.
Cumulative Concussion Force
Besides the risk of an individual blow to the head, there is a real and more recently known threat of long term cumulative head trauma. As this is still something being looked into the method of reporting this is not currently clear. One method may be to use a running average to determine the average force sustained by the athlete’s brain. Another method may be some form of average but were higher and potently more recent traumas hold a higher influence of the final value. More research into this would need to be conducted to make this value a usable and meaningful number.
Monitoring and Reporting
The data obtained by the onboard sensors would be relayed to a coach’s Wi-Fi or Bluetooth enabled laptop for real time updates. There would also be alerts for urgent warning such as heart abnormalities or extreme heat trauma or body temperatures.
This data would be transmitted using an Anaren Air Bluetooth radio or a Texas Instruments CC3100 SimpleLink Wi-Fi uC. For simplicity the data would be uploaded to the cloud were the couch could log in and see his teams stats on a single or possible multiple pages. The last possibility would to use a simple radio (eRIC by LPRS) and send the data to a PC and have a native program running that would interpret the various data coming in and display it on the coach’s PC.
Due to the shear size and complexity of what I have proposed it would be easy to not be able to deliver what I have mentioned above. There is also the challenge that working alone with only the input of some experts (cardiologist, biological signals engineer and potentially a neurologist) I may not be able to fully grasp what needs to be done to fully predict each symptom or how to correctly integrate the various sensors to get a correct a meaningful measurement.
To help resolve some of these issues I have outlined below the minimum deliverables I would like to have so as to have a useful project while still keeping the project to a manageable size. Besides this I have started looking at online help (courses and books) to help better understand signal processing and filtering as well as advanced algorithm design. Lastly were possible I will reuse open source libraries (Kalman filters, etc.) and input from friends and co-workers to help reduce the need to learn from costly and timely mistakes.
To keep the project at an achievable level the deliverables I propose will be somewhat less than those mentioned above.
- Heat stroke will be fully implemented looking for abnormal body temperatures.
- Heart attack will be implemented on a simple level; this would be only looking at abnormally high heart rate or abnormally high climb in heart rate. There is the potential to look for an abnormal heart rhythm time permitting. However, learning the athletes unique heart rhythm will most likely not be a possibility due to time constraints and potential complexity in doing such.
- Individual concussion severity will be fully implemented and have a high degree of accuracy this should use some form of filtering to ensure only actual concussions are registered and measured
- Cumulative concussion force may not be fully or correctly implemented. As it is still not fully clear in the scientific community how this should be measured, my attempts to do so may appear flimsy and illogical. While there may be an attempt at this it may be incorrect and inaccurate and therefore left out in the final products.
- Monitoring and reporting, for this the CC3100 in conjunction with a MSP430F5529 will be used to post the data to a cloud based system such as Plot.ly to give real time information on the athletes sensor measurements.
In closing I understand that this is a rather large project and will try my best to accomplish as much as possible. This is a project I am passionate about due to its ability to help athletes in the short term and potentially many more people in the long term. I believe that even if what I do does not directly get used it will have the potential to be used or inspire others to what other approaches are available in this interesting and exciting space.