Ice Hockey is a fast paced and high energy game. It can’t really be compared to any other team sport. Field hockey might come close because it is played with a stick controlling a ball and the aim is to score as many goals against an opposing team. But the one difference of the game environment changes everything – ice vs field. Ice hockey Players need to skate instead of run; besides trying to move fast with their hockey stick, they need to stay upright and have good control. Having good control might be easier if a person is skating alone; add in 9 other people zooming around in the 61 by 30m ice rink, moving at speeds upwards of 32km/h, collisions (whether intentional or not) is a common sight. So protective gear including helmets is a neccessary part of the game.
And due to the high-speed, accelerating and decelerating nature, players can’t sustain that level for long and they get substituted pretty often to maintain the high level of play. From that lens, it can be pretty challenging to keep track of how hard each player was going and how long they should rest/recover before heading out again. An athlete monitoring system will be handy in this case as it could track when each player was out or on the bench, how fast they were going and how many high-intensity accelerations they did as the game progressed. If positions of the players need to be tracked, the monitoring system will be very different compared to the ones used in football. That’s because most football games (whichever code of football) are played outdoor and position tracking can be done using GPS sensors. But since ice hockey is played indoor, GPS tracking wouldn’t work.
Then when we consider the skating component of the game and controlling the puck with the stick, there are opportunities to track the different movements of skating using sensors placed on the skate equipment. This could be useful for monitoring biomechanics of players who returned from injury or it could be a more sophisticated way of automatically identifying different activities on the rink (eg accelerating, passing, different shot types, tackling etc).
We wanted to have a look at some technologies that help manage two aspects of the sport – 1) injury prevention/monitoring and 2) performance monitoring. Realising this might make too long a post, I decided to just focus on the technologies that help prevent, monitor and manage injuries in this post and perhaps cover performance monitoring in another post.
Head Injuries & Concussions
Injuries in Ice Hockey can be mainly due to impact or fatigue. Injuries due to impact can be mitigated with proper protective gear while injuries due to fatigue can be managed with proper physical prep (strength and conditioning), rest (sleep), nutrition and hydration. [For those who would like to dive deeper into the details of Ice Hockey injuries and treatment from a sports medicine perspective, check out this paper – link]. With impact-related injuries, one of the top things that come to mind is preventing and managing concussions. Concussions can be caused by a direct impact on the athlete’s head or torso, or an indirect hit to the brain caused by sudden deceleration. The best approach to preventing head injuries and concussions is for an athlete to have hyper-awareness (think spidey-senses) and the ability to avoid impacts. But since that is not available, the next best approach to protecting against head injuries is using helmets.
Helmets, especially, full-faced masks helmets can be quite effective at protecting an athlete’s head and face. On the other hand, there is inconclusive evidence that they are effective against concussions in Ice Hockey. This could be partly due to a lack of published studies in that area. Nevertheless, companies and research institutes have been continuing research and development (R&D) to better the designs of helmets. Most of them do it with the aim of reducing linear and rotational accelerations of the head. Companies have looked into better materials, better helmet constructions and better fit designs. Some have even looked into better ways of testing and evaluating the helmets. Virginia Tech’s Helmet Lab developed a STAR Evaluation system and evaluated 55 different ice hockey helmets ranking them based on their ability to reduce concussion risks.
At the top of that list is CCM’s FL500. CCM having developed Ice Hockey equipment for many years, have also invested lots into R&D to make their helmets safer. Here are some of their innovations within their range of helmet designs:
- R.E.D system – which stands for Rotational Energy Dampening system consists of liquid-filled bladders incorporated into the liner. These bladders help manage/reduce the rotational accelerations during impact by creating a decoupling between the shell and the liner.
- Impact Pods – at one point, CCM had these Impact Pods which is a moulded shock absorber using expanded polypropylene (EPP). They have a specially engineered structure inside that absorbs energy from linear accelerations.
- D3O Smartfoam – CCM also partnered with D3O to incorporate D3O’s engineered shock-absorbing material together with their liner. The unique property of this material is it’s soft and flexible in a natural state but on impact, the material’s molecules lock together to dissipate impact energy.
- Microdial – is a feature that allows a 360 deg wrap and adjustment/customisation so that the helmet fits optimally on the athlete’s head. This is critical as research has shown that better fitting helmets reduce the risks of traumatic brain injury.
Similar to the technologies that CCM has incorporated in their helmet designs, the other helmets in the top 5 (including Bauer and Warrior) all have some form of innovation that involves impact-absorbing materials or structures. They also made sure to have a design that allows a good fit on the athlete’s head.
Head Impact Monitoring With Wearable Sensors
Besides trying to reduce the risk of concussions (using helmets), another strategy is to measure and monitor impacts using wearable sensors. The idea is that if an athlete has an impact during a game, and it was quantified, it can help the coaching team to make better-informed decisions; i.e. let the athlete keep playing or bench the athlete for the game or go for further diagnostics. A look into various research papers that investigated impact sensors and sports-related concussions generally gives us 2 main takeaways: 1) you cannot rely on impact sensor systems for real-time concussion screening and they should not replace clinical judgement, 2) Impact sensors could provide sideline staff with estimates of athlete exposure but not all sensor systems work the same way or have reliable/accurate estimates.
There has been close to 20 different head-impact sensor products that were launched or almost launched in the market in the last 15 years or so and they largely fall into three categories: headwear-mounted, skin patches and mouthguards. Most of them use inertial sensors or IMUs and a threshold and/or algorithm to process the sensor data and determine if a head impact has occurred. We won’t go through all of them (sensor products) but will talk about some of the key things in each category.
Headwear-mounted sensors are either mounted on/in helmets, embedded into a skull cap or secured on headbands. They were probably the earliest type to enter the field especially the ones that were built in/on helmets. One challenge with these type of sensors is they need to be fitted very well onto the athlete’s head so that the acceleration captured is close to what the skull is experiencing. If the helmet is loose, the sensor could be capturing additional movements and causing ‘false positives’. At least one study has shown that they do overestimate accelerations. Another study looked at the skull cap sensor and found that it performed very inconsistently. I think the concept of having sensors on the helmets or worn close to the skull is a great idea both in terms of operationally managing a set of sensors for a team of athletes and being unobtrusive to the athlete. Unfortunately, they miss the mark on accuracy and consistency. That could be part of the reason why most of them are no longer available.
Skin patches or skin-mounted sensors are sensor patches that can be stuck to the skin. The X-Patch by X2 Biosystems is possibly the only one of its kind. They are meant to be stuck on an athlete’s neck behind the ear. The logic is great because the placement location is close to the skull and it can be used by athletes of any contact sport regardless of whether a helmet is worn or not. There were quite a number of studies that investigated its usefulness both in the laboratory and on the field. The same study that looked at the skull cap sensor also found that the X-Patch had huge errors. A systematic review noted that the device accuracy is affected by skin motion and it also pointed to other studies that saw high measurement errors and found high numbers of ‘false-positives’ when compared to video captures. Another study that was conducted in the lab had results that showed the X-Patch overestimating peak linear acceleration. So it is not surprising that they are no longer selling it.
Mouthguard sensors, as the term suggests are mouth guards with sensors embedded in them. A Mouthguard sensor, when it is clamped in the jaw, is in indirect contact with the skull and is able to estimate the acceleration experienced by the skull. An advantage of having sensors in mouth guards is: they are already worn by athletes to protect their teeth. So having a mouthguard sensor doesn’t add an extra piece of equipment to the athlete. However, a couple of things to note about Mouthguard sensors are – 1) Mouthguard sensors with the sensors placed in a protruding tab can experience mechanical resonance and so have errors in peak acceleration measurements (reference); 2) If the Mouthguard sensor is not clamped properly or if the Mouthguard sensor is not a good fit, the measures will not be a good representation of head/skull movement alone. That said, among the three types of sensors, Mouthguard sensors seems to be still active in the market, unlike the others.
Concussion Assessment And Management Tools
Impact sensors as we have seen above do have errors and have a chance of giving ‘false positives’. So another step in dealing with possible concussions is to use a concussion assessment tool. A popular tool is the SCAT or Sports Concussion Assessment Tool currently in its 5th edition. It is designed to be used only by physicians and licensed healthcare professionals. It can also be used as a preseason baseline test which helps to interpret post-injury test results. If there are no licensed healthcare professionals, the coaching team can use the CRT or Concussion Recognition Tool which is still helpful for identifying suspected concussions.
A number of Apps/programs have been developed that incorporate similar assessments to help with recording and managing a team of athletes. They typically have some/most of these features: 1) preseason baseline testing, 2) post-injury assessment, 3) return to regular activity procedure, 4) post-season testing/assessment, 5) communication/reporting tools, 6) support from medical or allied health professionals, and 7) educational resources & tools. Some of the Apps that provide a structured assessment include CSx, HeadSmart, easySCAT, CARESport and HeadCheck.
Objective Diagnostic Assessments
One shortcoming of SCAT/CRT or any other similar assessments that are reliant on the athletes’ subjective response is that they can be manipulated if the athletes are determined to avoid a concussion diagnosis and keep playing. Thus objective diagnostic measures can provide a more accurate concussion evaluation. We came across two objective diagnostics in a paper (link) and they are:
- King-Devick Test: The KD Test is a vision-based performance measure. During the test, Athletes have to read out loud pages of single-digit numbers shown on an iPad and they are evaluated on accuracy and timing. Baseline measures are done and used to compare with subsequent tests after suspected concussions. Research has shown that it can detect concussion with high levels of sensitivity (86%) and specificity (90%) at rinkside. Here’s a video that better demonstrates its use: link.
- Quantified EEG using NeuroCatch: NeuroCatch is a portable hardware and software system that is worn on an athlete’s head and measures brain activities. The system processes the EEG data and derives Event-related potentials (ERP) which reveals neurofunctional deficit. Basically, I think what it means is when (again) compared to a baseline measure, it can determine concussion effects and also assess recovery post-injury and rehab.
Shoulder & Neck Injuries
Aside from concussion, shoulder dislocations and injuries to the neck (comparable to a whiplash) are pretty common in Ice Hockey. These are also often due to impacts and (again) other than trying to prevent those impacts, one way to mitigate those injuries is with really good protective gear.
As mentioned earlier in the helmets section, D3O, a protective products company, has developed innovative materials that have high shock-absorbing properties. Using unique patented and proprietary technologies, the formulated materials are soft and flexible in its natural state, but on impact, literally ‘toughens up’ to dissipate impact energy and so protecting the wearer. If you haven’t seen how the material responds to impact, you should check out the demo in this video: youtube link.
D3O helped to develop CCM’s shoulder pads with the material inserted in the shoulder caps, sternum and spine. They also worked with Aegis Impact Protection to develop their Interceptor Bib, a neck guard with D3O material insert to make it impact and slash-resistant. Are there other materials that can absorb impact as well? Yes. A study tested and compared 5 different materials to see their effectiveness in absorbing impact. At 5mm of material thickness, D3O fared the best, followed by a material called Gphlex (which isn’t around anymore, unfortunately) then Poron XRD, and leather and lastly EVA foam (e.g. yoga mat). But if thickness (or bulkiness) is not a concern, 15-20mm thick EVA foam could also absorb impact really well. Although athletes will then have to compromise on agility and flexibility.
Wearables/Technologies For Rehab
But life is such, accidents still happen and injuries occur and when injuries are bad enough, the recovery process will include rehabilitation. There are various technologies that can come to play in this area whether its in terms of assessing range of motion or ROM at the Physiotherapist’s, tracking day-to-day exercises or monitoring posture.
Assessing ROM: There are non-wearable motion capture systems that provides highly accurate tracking but they tend to require elaborate setup, calibration and preparation and not to mention they can be expensive. Wearable sensors/systems on the other hand may not be as precise and they also require calibration but they can be quite easy to setup and use, and (quite importantly) they cost a lot less. There are a number of off-the-shelf IMUs that are capable of doing that; they may not neccesarily have the exact application for measuring ROM but they provide (developer) tools to build custom applications. A few examples include Mbientlab, YOST labs – 3 Space Sensors , and Movesense. We did come across a couple of products with applications designed for assessing ROM. One is Lynx by DyCare. The Lynx system uses IMU sensors developed by Shimmer and by using two IMUs, it is able to measure joint angle in real-time plus other movement metrics. Another one is Sens by K-invent. Interestingly, it tries to do the same job of measuring range of motion and joint angle using only one IMU. It seems like it works by setting the starting position as ‘zero’ and calculates angle changes as the user moves the limb to the final position. But I think that relies on the user being very stable and only moving that specific limb, else it will affect the measurement accuracy.
Tracking/Monitoring Exercises: Usually, after the initial assessment, the athlete will be prescribed exercises to do that will aid recovery. There are a couple of sensors are that about to be launched in the market that can track exercises on exercise bands (which are typically used for rehab). There’s the Prohab Smart Gauge which is a sensor that can be attached to the end of an exercise band or cable machines. It can track in real-time, the effort of every rep and how fast each rep was completed. This could provide good feedback to the Physio or Physical Therapist of how well the athlete is recovering between visits. A similar product is STRAFFR which is an exercise band with in-built sensors. Although it is marketed more as a personal training device, it can potentially be used as a rehab tracking tool. As mentioned, they are about to be launched and can be pre-ordered on their websites.
Posture Monitoring: Posture plays a big part in muscle and injury recovery especially where shoulders and neck are concerned. Similar to assessing range of motion, posture monitoring can also be achieved using IMUs and all the above mentioned off-the-shelf IMUs have the potential to be used as posture monitoring devices if an application was developed for them. But one product in the market that was developed for this purpose is the Lumo Lift. It is a sensor device designed to be ‘stuck’ on the inside of your shirt right below the collarbone. It is secured in place on the inside of the shirt using a small magnet piece placed on the outside of the shirt. The device needs to be ‘calibrated’ for the correct/targeted posture and when it detects that the user is slouching, it vibrates reminding the user to get back to the right posture.
Technologies From Startups
We just had a look at three common injuries in Ice Hockey – head, shoulder and neck injuries. There’s quite a lot of technologies and innovations that are out there aiming to cope with those injuries – from gear to protect (from those injuries) to technologies that measure the magnitude of impacts (mainly for head injuries), and diagnostic tools and technologies to help with rehabilitation. As always, startups tend to be at the front line of coming up with new solutions. At the same time, because of different reasons they don’t quite get to the end of the tunnel (especially with limited funds). Those that have held on and refined their solutions and gotten extra funding and/or found the right partnerships and made it, should really be applauded.
Opportunities To Explore
There is a study that evaluated collisions from 54 games and they identified the different collisions types, the level of anticipation and body position prior to collision. It was found that 63% of the collisions were along the playing boards and the remaining happened on open ice. Their findings also suggest that with heightened player anticipation of impending body contact, the head impact severity decreases. These findings show that it would be beneficial to focus on training the athletes’ awareness or ability to anticipate and deal with collisions. To make it safe, this could be an opportunity to adopt virtual reality (VR) training (like Sense Arena) and incorporate collision anticipation training. Also, with more than half the collisions taking place along the playing boards, better impact absorbing materials or structures should be explored if it hasn’t been yet.
Of course, apart from technology interventions, improved regulations and greater awareness of injuries (especially concussions) and good physical preparations of the athletes (including hydration) can also go a long long way in injury prevention.
I hope you all found some value in this piece. Thanks to Stuart who inspired me to get this going and gave me some suggestions. The technologies mentioned above definitely doesn’t cover everything that is out there. I am sure there are some that I missed out. Please feel free to leave a comment or feedback below or drop me a message here (link). With that, thanks for reading!