Concussions, classified as mild traumatic brain injuries, are one of the most common forms of injury in the United States, with an estimated 1.6 to 3.8 million people experiencing a single concussion each year, with around half of cases going unreported or undetected (American Brain Foundation). Even with such high numbers of incidence, there is still no universal objective diagnostic tool, leaving diagnosis up to medical professionals' subjective discretion. This gap has opened a new world of research, with many scientists, physicians, and other medical professionals looking for new ways to diagnose concussions objectively.
Earlier this year I had the privilege of attending a neuroscience seminar where a very promising biomarker was identified as being useful in the diagnosis of concussions. Dr. Jennifer Krizman, a neuroscientist at Northwestern University, along with some of her colleagues, conducted a partially longitudinal, cross-sectional study that followed 40 children, 20 of whom had been diagnosed with concussions, and found that their neural processing of sound was disrupted following their injuries, making the neural coding of the fundamental frequency (F0) a possible biomarker for concussion. More specifically, an affected individual’s ability to perceive speech in noise is based off the brain’s F0 encoding ability. Therefore, disruptions of this process result in a pattern change that can be measured and compared to non-concussed individuals.
Similarly, in a news article published by New York Presbyterian, a different biomarker has been identified as providing a possible measurable objective indicator of concussions. While still in the developmental process, Dr. Thomas Bottiglieri, a physician at Columbia University Irving Medical Center, proposes that concussions reflect disruptions in proprioceptive neural circuits, which can be measured through the use of machine learning analysis. The tool, called ProScope, is reported to detect the biomarker with 80% to 90% sensitivity through the use of AI-based pattern recognition algorithms that are able to identify subtle physiological changes that are almost impossible to detect through traditional clinical assessments.
Both Dr. Bottilgieri’s and Dr. Krizman’s biomarker findings, while focused on different brain regions and functions, are based on the fundamental premise that concussions disrupt the brain’s normal processes, affecting cognition and producing measurable physiological changes. All brain processes happen in a pattern-like nature; therefore, no matter what brain region or ability is impaired, if the pattern is disrupted, an AI system can measure and identify these changes with way more efficacy than human healthcare providers.
It is interesting to think about the possibilities that the use of AI could bring to healthcare; in this case, we are one step closer to an objective diagnostic tool for one of the world’s most common injuries.
No comments:
Post a Comment