Concussions are a common occurrence in sports. The issue that clinicians and others have when deciding when an athlete can return to play after a brain injury is similar.
Most athletes recover from a sports-related concussion in seven to ten days, but some take longer. Because of this situation, treating sports-related concussions is extremely difficult.
A innovative approach has been developed by researchers from Florida Atlantic University’s College of Engineering and Computer Science and SIVOTEC Analytics in Boca Raton, as well as colleagues. They’re teaching robots how to anticipate concussion recovery time based on symptoms like headache, dizziness, and weariness.
Their research, which was published in the journal Medicine & Science in Sports & Exercise by the American College of Sports Medicine, can be utilized to build a decision support system that will help clinicians establish tailored treatment plans for injured athletes. This study is part of a bigger effort by the researchers to develop machine learning models to aid in the diagnosis, tracking, and treatment of a variety of brain health conditions.
The researchers evaluated data from 2,004 concussion events in 22 sports using data from the National Athletic Treatment, Injury, and Outcomes Network (NATION), an injury monitoring program for high school student-athletes, to see where the injuries predominantly occurred. They discovered that American football was responsible for more than half of the concussions.
They built a new dataset of concussion injuries in football, as well as other contact sports such as wrestling, field hockey, boys’ and girls’ basketball, soccer, and lacrosse, using this information.
There were 922 football concussions and 689 concussions from other contact sports in this new dataset, for a total of 1,611 concussion events across all contact sports.
We have introduced a cutting-edge approach and new clinical tool to manage sports-related concussions, which will measurably improve with more and more inclusive data. Our supervised machine learning method has demonstrated efficacy and warrants further exploration.
Taghi Khoshgoftaar
The total number of symptoms reported per sports-related concussion episode ranged from zero to seventeen in the dataset of all contact sports, with 55 percent of student-athletes reporting five or more symptoms. The researchers used a supervised machine learning-based modeling strategy to predict concussion-related symptom recovery times in seven, fourteen, and twenty-eight days.
They used a dataset containing three years of concussions suffered by these high school student-athletes in football and other contact sports to test the performance of ten categorization algorithms in creating prediction models.
The symptom-based prediction models demonstrated practical clinical value in estimating sport-related concussion recovery time, with the most common reported sports-related concussion symptom being a headache (94.9 percent), followed by dizziness (74.3 percent), and then difficulty concentrating (61.1 percent) in the dataset.
In the case of concussion case management and patient care, this knowledge can be extremely useful to health care workers. This knowledge can aid with academic accommodations and team needs in addition to clinical decision-making.
“We have introduced a cutting-edge approach and new clinical tool to manage sports-related concussions, which will measurably improve with more and more inclusive data,” said Taghi Khoshgoftaar, Ph.D., co-author and Motorola professor in FAU’s Department of Computer and Electrical Engineering and Computer Science, who collaborated with lead author Michael F. Bergeron, Ph.D., senior vice president of development and applications at SIVOTEC Analytics, and Sara Landset, co-author and a Ph.D. student at FAU. “Our supervised machine learning method has demonstrated efficacy and warrants further exploration.
The researchers found that the overall number of symptoms, sensitivity to noise or light, difficulty concentrating, insomnia, and balance concerns have the highest predictive value, implying that they play a significant role and are useful in their models.
Amnesia, hyperexcitability, loss of awareness, and tinnitus, on the other hand, were not shown to be suitable possibilities for quantitatively assisting top-performing models.
“It is really important to be able to promptly identify those athletes who are going to need more time to recover after incurring their concussion,” said Bergeron.
“The ability to predict recovery time using machine learning will help to augment an effective stratified approach to care. This also can help with realistic expectations of the student-athlete, as well as provide important insight and perspective for parents, coaches and teachers.”
Collaborators on the study, “Machine Learning in Modeling High School Sport Concussion Symptom Resolve,” are Nemours Children’s Hospital, Division of Neurosurgery in Orlando; Cedars-Sinai Kerlan-Jobe Center for Sports Neurology in Los Angeles; and Datalys Center for Sports Injury Research and Prevention, Inc. in Indianapolis.
“This novel application of supervised machine learning to sport concussion epidemiology is an important step in advancing the approach in clinically managing a complex condition,” said Stella Batalama, Ph.D., dean of FAU’s College of Engineering and Computer Science.
“Supervised machine learning has the potential to more effectively reveal meaningful patterns and potentially unique vital insights into the complex inter-dependent array of clinical determinants in anticipating concussion symptom recovery as well as myriad other aspects in managing concussions.”
