Determination of Fall Risk for Lower Limb Amputees
Abstract
Falling is a common problem for lower limb amputees. More than half of all lower limb amputees report falling at least once in the past year. Of those that do report a fall, about 75% reported multiple falls. Falling can have multiple negative results including broken bones, traumatic brain injuries, cuts, bruises, sprains, or even death. Most times a fall results in minor injuries such as scrapes and bruises, but the fall can impact a person’s confidence in their balance and mobility. This may lead a person to reduce their physical activity and interaction with friends and family, leading to a reduction in physical and emotional health. While it is known that lower limb amputees do fall, the specific number of falls in the Military and Veterans Affairs (VA) amputee communities is not well understood. This is because evaluations of falling are often based on an individual’s memory over an extended period. This can be particularly true for the Military and Veteran lower limb amputee. At some point, 62% of Military and Veteran traumatic amputees have experienced a traumatic brain injury that puts them at increased risk for falling and can affect their ability to remember those falls accurately. Furthermore, 86% of Veteran amputations are due to diabetes and peripheral vascular disease, and the majority are in the 65-74 year-old age range, both creating an elevated risk for falling. An improved understanding of falls could lead to improved rehabilitation care. The purpose of this study is to create and test an advanced sensor package that can be placed on the prosthesis of a lower limb amputee that will identify when falls occur. The proposed research has two aims: (1) establish a baseline fall detection algorithm derived from simulated falls in a laboratory setting and (2) utilize and refine the initial laboratory-based algorithm to provide detection of fall events during activities of daily living in pragmatic, real-world environments. To achieve these aims, we will perform two human subject experiments. The first experiment will use 30 non-amputee and 5 lower limb amputee individuals to simulate falls in a laboratory setting while wearing the sensor. The sensor will record the motion of the body while falling so that we can create an algorithm to detect a fall in comparison to normal daily activities. The second experiment will recruit 40 lower limb amputees to wear the sensor in the real world. Amputees will use the sensor for an 8-week period. During that time, the sensor will record their motion and detect when a fall occurs. Participants will also report weekly about any fall events that were not detected so that the algorithm can be improved. The outcomes from this 2-year project will be new information for clinicians to better understand the number of falls that occur for lower limb amputees. This work represents an initial pilot study to collect data for the fall detection algorithm and lead to future studies where large numbers of amputees will be supplied with the sensors in order to better quantify falling in the larger amputee community and other communities that are at high risk for falling. The long-term goal of this project is to create a sensor that is accurate enough to detect falls with minimal errors and report them to the clinician so that updates to rehabilitation care can be assessed prior to a planned visit. In addition, the ability to sense and communicate a falling event in real time will allow the appropriate medical personnel to be notified more quickly. This would allow for a faster emergency response for individuals in the event of an injurious fall and reduce time for medical care.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 10, 2021
- Source ID
- W81XWH2010164
Entities
People
- Richard R Neptune
Organizations
- United States Army
- University of Texas at Austin