A breakthrough AI tool has been developed to identify subtle structural changes in the brains of athletes caused by repeated head injuries, offering improved diagnosis and understanding of such injuries over time.
A recent study involving student-athletes has revealed that an artificial intelligence computer program, specialized in analyzing magnetic resonance imaging (MRI), can effectively detect structural alterations in the brain resulting from recurring head injuries. Previously, these changes went unnoticed with traditional medical imaging techniques like computerized tomography (CT) scans. This groundbreaking technology has the potential to contribute to the development of new diagnostic tools for a better comprehension of subtle brain injuries that accumulate over time.
The potential risks of concussions among young athletes, especially those engaged in high-contact sports like football, hockey, and soccer, have long been recognized by experts. There is increasing evidence that repeated head impacts, even if initially considered mild, can accumulate over many years and lead to cognitive impairment. While advanced MRI can identify microscopic changes in brain structure resulting from head trauma, researchers point out that the extensive data generated by these scans is difficult to navigate.
Led by researchers in the Department of Radiology at NYU Grossman School of Medicine, the study demonstrated for the first time that an innovative AI tool, utilizing machine learning, can accurately differentiate between the brains of male athletes who played contact sports such as football and those engaged in non-contact sports like track and field. The results established a connection between repeated head impacts and subtle structural changes in the brains of contact-sport athletes who had not been diagnosed with a concussion.
“By comparing the brains of athletes in contact sports to those in non-contact sports, we have identified significant differences that indicate potential risks associated with certain sports,” said Dr. Yvonne Lui, the senior author of the study and a neuroradiologist. Dr. Lui, who is a professor and vice chair for research in the Department of Radiology at NYU Langone Health, added, “These findings suggest that there may be a risk in choosing one sport over another, even if we expect them to have similar brain structures.”
In addition to its potential for detecting damage, the machine learning technique employed in this investigation may contribute to a better understanding of the underlying mechanisms of brain injuries.
The study, recently published in The Neuroradiology Journal, involved hundreds of brain images from 36 college athletes engaged in contact sports (mainly football players) and 45 athletes participating in non-contact sports (mostly runners and baseball players). The objective was to establish a clear link between changes detected by the AI tool in the brain scans of football players and head impacts. This research builds upon a previous study that identified differences in brain structure between football players, both with and without concussions, and athletes participating in non-contact sports.
For the investigation, the researchers analyzed MRI scans taken between 2016 and 2018 from 81 male athletes who had no known diagnosis of concussion during that period. Contact-sport athletes participated in football, lacrosse, and soccer, while non-contact-sport athletes engaged in baseball, basketball, track and field, and cross-country.
During their analysis, the research team developed statistical techniques that enabled their computer program to “learn” how to predict exposure to repeated head impacts using mathematical models. The program became increasingly adept as more training data was provided.
The program was trained to identify abnormal features in brain tissue and distinguish between athletes with and without repeated head injuries based on these factors. The researchers also assessed the usefulness of each feature in detecting damage, helping to identify which MRI metrics contributed most to the diagnosis.
According to the authors, two metrics accurately identified structural changes resulting from head injuries. The first metric, mean diffusivity, measures the ease of water movement through brain tissue and is often used to detect strokes in MRI scans. The second metric, mean kurtosis, evaluates the complexity of brain tissue structure and can indicate changes in brain regions involved in learning, memory, and emotions.
“Our results demonstrate the power of artificial intelligence in revealing previously invisible injuries that do not appear on conventional MRI scans,” said Junbo Chen, the lead author of the study and a doctoral candidate at NYU Tandon School of Engineering. Chen added, “This method may offer an essential diagnostic tool not only for concussions but also for detecting damage resulting from subtler and more frequent head impacts.”
Chen also mentioned that the research team plans to explore the application of their machine-learning technique to examine head injuries in female athletes.
The study was funded by the National Institutes of Health and the U.S. Department of Defense.
Reference: “Identifying relevant diffusion MRI microstructure biomarkers relating to exposure to repeated head impacts in contact sport athletes” by Junbo Chen, Sohae Chung, Tianhao Li, Els Fieremans, Dmitry S. Novikov, Yao Wang, and Yvonne W. Lui, 22 May 2023, The Neuroradiology Journal.
By clicking “Yes,” you consent to the use of Google Analytics and related cookies across the TrendMD network (widget, website, blog). Learn more
By clicking “No,” you do not consent to the use of Google Analytics and related cookies across the TrendMD network (widget, website, blog). Learn more
Frequently Asked Questions (FAQs) about brain injury detection
What is the purpose of the AI tool mentioned in the text?
The purpose of the AI tool is to detect subtle structural changes in the brain caused by repeated head injuries in college athletes.
How does the AI tool work?
The AI tool utilizes machine learning and is adept at processing magnetic resonance imaging (MRI) scans. It analyzes brain images and identifies abnormal features in brain tissue associated with head injuries.
How is the AI tool different from conventional medical imaging methods?
Unlike conventional medical imaging methods like computerized tomography (CT) scans, the AI tool can detect structural changes in the brain that were previously undetected. It can identify microscopic changes in brain structure resulting from head trauma, providing more comprehensive information.
What are the potential benefits of the AI tool?
The AI tool has the potential to improve diagnosis and understanding of subtle brain injuries that accumulate over time. It may aid in the development of new diagnostic tools and contribute to a better understanding of the underlying mechanisms of brain injury.
What did the study involving student-athletes reveal?
The study demonstrated that the AI tool could accurately distinguish between the brains of male athletes who played contact sports (e.g., football) and non-contact sports (e.g., track and field). It linked repeated head impacts to tiny structural changes in the brains of contact-sport athletes who had not been diagnosed with a concussion.
How many athletes were involved in the study?
The study involved hundreds of brain images from 36 contact-sport college athletes (mostly football players) and 45 non-contact-sport college athletes (mostly runners and baseball players).
What were the key metrics used by the AI tool?
The AI tool utilized two key metrics: mean diffusivity and mean kurtosis. Mean diffusivity measures the ease of water movement through brain tissue, while mean kurtosis evaluates the complexity of brain tissue structure.
Was the study funded by any organizations?
Yes, the study was funded by the National Institutes of Health and the U.S. Department of Defense.
More about brain injury detection
- AI Tool May Help Spot “Invisible” Brain Damage in College Athletes
- Study: AI Tool Detects Subtle Brain Structure Changes in Athletes
- The Neuroradiology Journal – Identifying Relevant Diffusion MRI Microstructure Biomarkers
- NYU Grossman School of Medicine – Department of Radiology
- National Institutes of Health
- U.S. Department of Defense