You are viewing 1 of your 1 free articles
Watch out! The age of artificial intelligence (AI) is here – the global ecosystem is rapidly evolving, and the sports therapy industry needs to evolve along with it! Jason Tee provides clinicians with a cheat sheet for how to incorporate the power of AI in clinical practice to ensure that they don’t get left behind.
Auburn Tigers quarterback Payton Thorne is tackled by Texas A&M Aggies defensive back Bryce Anderson and linebacker Taurean York during the first quarter at Kyle Field. Mandatory Credit: Maria Lysaker-USA TODAY Sports
Artificial intelligence (AI) is an innovative technology that utilizes machine learning, neural networks, and expert systems algorithms to simulate human intelligence, learn from data, and accomplish complex tasks. ChatGPT is the poster child for what conversational AI language models can do. Still, many rapidly developing competitor chatbots, such as Amazon’s Alexa, Google Assistant, and Microsoft’s Bing, are in the market. OpenAI developed ChatGPT, a powerful deep-learning model that can understand and generate human-like responses. ChatGPT and similar bots provide information, answer questions, and assist users across various platforms, revolutionizing how they interact with technology and enhancing digital experiences. These sophisticated language models are trained on massive amounts of text data, enabling them to engage in natural and interactive conversations. The critical difference between AI language models and previous-generation search engines like Google is that search engines would find web pages and return individual websites as solutions. Artificial intelligence language models integrate and synthesize from multiple information sources, generating unique but informed responses*.
*Disclaimer – ChatGPT wrote the above paragraph in response to the prompt “Write a one paragraph summary on what AI is for a popular media publication. Include an explanation of how ChatGPT works. Include names of other AI chatbot products.”
Artificial intelligence language models can potentially democratize access to information and expertise. They can answer questions instantly, offer suggestions, and assist with problem-solving. This empowers individuals to access knowledge and resources that may have previously been limited or inaccessible. This advancement will profoundly affect expertise-based economies such as the medical field. Inevitably, potential clients will seek advice from chatbots before (or instead of) consulting with industry professionals. Sports medicine professionals must understand the capabilities and limitations of AI language models to adapt to these changes and explore new opportunities.
Artificial intelligence will undoubtedly affect how clients interact with their therapists and medical support team—got tight hip flexors? Weak hamstrings? Struggling with back pain or knee pain? Clients can ask ChatGPT to generate workouts designed for them. The depth of what ChatGPT can produce is impressive. Look at the response below for a request for an injury prevention program for a girls’ football team.
ChatGPT Prompt |
I am a football coach for a U15 girls’ team. Last year, we had a lot of ankle and knee injuries. Generate a 10-minute program that we can use during warm-ups to prevent these injuries. Include specific exercises, sets and reps, and three levels of progression. |
Response |
Warm-Up Injury Prevention Program for U15 Girls Football Team (10 minutes)
Level one (Beginner):
Level two (Intermediate):
Level 3 (Advanced):
Note: Ensuring proper technique and form during all exercises is essential. Progression to higher levels should only occur when players have demonstrated proficiency and are physically ready. Coaches should monitor players closely and adjust based on individual abilities and limitations. |
Similarly, AI is also able to guide injury diagnosis and rehabilitation. In 2021, a research group from the University of Potsdam demonstrated that given a series of sports injury case studies, an AI chatbot app could correctly diagnose, provide accurate advice, and refer patients to an appropriate medical specialist(1). In the same way, ChatGPT does a reasonable job of prescribing therapy and exercise programs for injury rehabilitation.
ChatGPT Prompt |
Provide an injury rehabilitation program for a male basketball athlete with a grade 2 MCL strain. Include exercise training prescriptions for each rehabilitation phase and suggested therapies at each stage. |
Response |
Please note that I’m an AI language model, not a healthcare professional. It’s important to consult with a qualified sports therapist or healthcare provider for an accurate and personalized injury rehabilitation program. Nevertheless, I can offer you a general guideline for a grade two MCL strain rehabilitation program:
Phase one: Acute Phase (0-72 hours)
Phase two: Subacute Phase (3-7 days)
Phase three: Intermediate Phase (1-4 weeks)
Phase four: Advanced Phase (4-8 weeks)
Remember, this is a general overview and should be adapted and personalized based on the athlete’s specific condition, progress, and guidance from a healthcare professional. Compliance with the prescribed program and regular communication with the medical team are crucial for successful rehabilitation. |
While it is notable that ChatGPT provides the necessary disclaimers and does indeed direct users to engage with the appropriate medical professionals, it is clear that medical professionals will no longer have the exclusive rights to knowledge on injury rehabilitation.
The World Economic Forum has estimated that artificial intelligence will replace some 85 million jobs by 2025(2). Fortunately for those working in medical and therapeutic industries, these are not projected to be unduly affected by AI in the short term (five years). The jobs most likely to be immediately replaced by AI are data entry clerks, administrative and executive secretaries, accounting bookkeeping, and payroll clerks.
Artificial intelligence will immediately replace task-oriented professions but will take far longer to infiltrate areas requiring professional judgment and expertise. The old saying goes, “Knowledge is knowing that a tomato is a fruit; wisdom is not putting it in a fruit salad.” Artificial intelligence applications can place knowledge at users’ fingertips. Still, the biggest limitation is that they lack the dynamic and interactive nature of a real-time conversation between a patient and a healthcare provider. Currently, AI cannot assimilate contextual information particular to your client to create bespoke, best-fit solutions. However, this challenges healthcare professionals to provide exceptional programs and services to differentiate themselves from AI applications because if they provide cookie-cutter advice, they are not providing much more than AI does.
Part of the expertise that clinicians bring to their work as medical specialists is the ability to ask good questions. The better the prompts that they can feed any language model, the more valuable the answers provided become. This is a critical differentiator between how clinicians might use AI versus a potential client.
Artificial intelligence applications have to be trained on a particular data set. In the case of ChatGPT, this was a data set known as the Common Crawl, which was current up until September 2021. Despite its widespread use, ChatGPT is still in testing and, as a result, does not have access to the most up-to-date information. However ultimately, the data set will be updated, but AI currently can’t keep up with those concerned with being evidence-based and using the most up-to-date research.
A further concern with AI is that it doesn’t follow the scientific method of knowledge generation and instead aggregates and assimilates information from a wide range of sources, including the Internet. Since anyone can publish on the Internet and not all information is accurate, AI language models are subject to misinformation. Similarly, because of what is available on the Internet, AI language models are subject to socioeconomic bias.
The introduction of AI to general public use has disrupted traditional work approaches. However, it’s not all bad, and for those willing to learn and adapt, there could be several benefits to their clients and business. Currently, the biggest development area for clinicians is the potential for improved productivity through assistance with administrative tasks. Business administrative tasks are time-consuming and preoccupy clinicians when they could be consulting with patients. Artificial intelligence offers various creative possibilities to enhance time management, information dissemination, and health outcomes. Clinicians should embrace the change and, through creative collaboration, find ways to include AI in clinical practice. Below are some ways AI will likely benefit sports injury rehabilitation and prevention.
Clinicians
Athletes
Our international team of qualified experts (see above) spend hours poring over scores of technical journals and medical papers that even the most interested professionals don't have time to read.
For 17 years, we've helped hard-working physiotherapists and sports professionals like you, overwhelmed by the vast amount of new research, bring science to their treatment. Sports Injury Bulletin is the ideal resource for practitioners too busy to cull through all the monthly journals to find meaningful and applicable studies.
*includes 3 coaching manuals
Get Inspired
All the latest techniques and approaches
Sports Injury Bulletin brings together a worldwide panel of experts – including physiotherapists, doctors, researchers and sports scientists. Together we deliver everything you need to help your clients avoid – or recover as quickly as possible from – injuries.
We strip away the scientific jargon and deliver you easy-to-follow training exercises, nutrition tips, psychological strategies and recovery programmes and exercises in plain English.