Posturography-Based Decision Support System for mTBI Detection
Abstract
A mild traumatic brain injury (mTBI) or concussion, as defined by the CDC, is the disruption of normal brain functions that can be caused by a bump, blow, or jolt to the head. Despite accounting for roughly 90% of all traumatic brain injuries, not all mTBIs are diagnosed or treated. If left untreated, mTBIs can lead to post-concussion syndrome and a number of long-term cognitive impairments. Another long-term effect of mTBIs is the inability to adequately control postural sway; trauma to brain centers responsible for maintaining balance and proprioception reduce the ability to stabilize when standing upright. Postural sway can be measured and assessed with posturography. Analysis of posturography results can provide insight into the physical condition of a patient’s brain following an mTBI or aid in the diagnosis of previously undetected mTBIs. In a clinical setting, however, it becomes challenging and time-consuming to conduct the structural analysis of posturography manually. This study leverages custom-made force plates to collect postural sway data from over 120 patients to create a decision support-system-like approach to the diagnosis of mTBIs using various classification models. With a high accuracy of 88%, this study examines a novel set of methodologies for mTBI detection that are mobile, inexpensive, and less time-consuming than traditional methods. The system resulting from this study has the potential to greatly increase the diagnosis rate of mTBIs, reducing the impact of untreated mTBIs. Future work includes the use of this system for the detection of neurodegenerative diseases, such as Alzheimer’s and Parkinsons.