Imagine this: Rugby players sprint down the pitch with their opponents invisible, just to collapse along the way. It is a contactless injury, annoyed, often preventable setback that allows players to be on the sidelines for weeks or months. Rugby is a game of power, accuracy and merciless strength. It is also a sport where injuries are always present.
But imagine tools that can be predicted before an injury occurs, giving the coach the opportunity to intervene and keep players in the game. This is the potential endpoint for the latest research into AI and rugby injuries.
Non-contact foot injuries account for almost 50% of players absent from Rugby Union players, often bystanding players for weeks or months in severe cases. These injuries, such as hamstrings, gro diameters, thighs and calf strains, can be extremely frustrating for both players and teams. It disrupts your training schedule and affects your choices and team performance.
Previous research was often lacking, and was lacking. They may have looked into how isolated factors such as age, previous injuries, and player flexibility are related to injuries, but they do not always consider the complex interactions between these factors. It's like trying to solve a puzzle by just looking at one at a time.
AI can now predict most contactless rugby injuries.
The reality is that older players with less flexibility in the joints returning from an injury, for example, have more flexibility and are at a higher risk of injury than older players who have not been injured in recent years.
Crack your code with AI
In our latest research, we took a different approach. Over 1,700 weekly data points were collected from full-time male rugby players over two seasons. These consist of factors known to be associated with non-contact leg injuries, such as changes in weight, training intensity, fitness parameters such as intensity and cardiovascular fitness, past injuries, and performance from muscle and joint screening tests. Before the training session, we saw how players felt at the start of each day.
We have fed this information to powerful AI systems that can find complex patterns. We sifted through all the data to find a combination of risk factors associated with players maintaining leg injuries.
The results were interesting. The AI model predicted severe non-contact leg damage with an accuracy of 82%. Therefore, every ten such times, the model would have correctly predicted eight.
This model suggested that players were at a higher risk of injury when there was a combination of reduced strength in the hamstring and gro diameter, reduced flexibility in the ankle joint, size of muscle pain, and frequent changes in training intensity.
This model predicted non-contact ankle sprain with 75% accuracy using other factors such as reduced sprint time, weight gain, and previous injuries and concussions. However, it also predicted several other less severe leg injuries with similar (74%) accuracy, but not all injuries are confidently predicted.
An early warning AI system can provide coaches with important insights that could put players at risk. Think of it as a high-tech crystal ball, it gives you a glimpse into the potential issues before they arise, allowing for aggressive measures to keep players on the field.
Coaches can use this information to create tailored training programs that ensure that players are monitored and supported consistently. Targeted interventions, such as exercises designed to address certain weaknesses or increase mobility, can significantly reduce the risk of injury.
In theory, by optimizing preseason training through focus athlete screening, our research may provide clear and practical guidelines. These simple and cost-effective tools allow coaches and medical staff to identify potential risks early and provide a proactive approach to player safety and performance.
This AI-driven approach is not just rugby. It can be used in any sport where data can be collected. Imagine a personalized training plan and injury prevention strategy for all athletes, from soccer players to gymnasts. It will change how athletes train and compete, helping them stay healthy and perform best.
AI is not yet widely used in elite sports. However, it is expected that the development of smart technology for watches that monitor training along with other factors will soon be possible to expand to recreational athletes.
The future of injury prevention?
However, this study is only the first step. Scientists around the world are already working on ways to make these AI models even more accurate by including other risks in athletes, such as psychological factors and indicators of physical movement. They also see how different sports have a unique combination of risk factors that need to be considered.
By combining AI accuracy with sports science and medical insights, we stand on the brink of revolution in injury prevention and performance optimization. This approach not only increases player safety, but could also make the most of their potential and redefine how athletes engage in the sport they love.
As a proven position in rugby, this innovation can pave the way for a safer and smarter future for sports.
Selen Evans is a physiotherapist and researcher at Bangor University in Wales.
Julian Owen is a senior lecturer in sports and exercise physiology at Bangor University, Wales
Reissued from the conversation
Published on June 15th, 2025 by EOS and Dawn
