Living in “My Home,” not in “A Home”
Every year, over 700,000 people suffer a stroke nation wide. Two-thirds of those who suffer a stroke survive and require rehabilitation. A weakening of the muscle, specific the leg muscles, is especially common. For the care provider, this can become burdensome, or even risky. This can lead to a condition where the stroke survivor must move into a care provided facility. An in-home harness system, which would support the stroke survivor and minimize of injury caused by helping to move the stroke survivor, could provide a means of continued living at home.
The goal of this project is to develop a harness support system for stroke survivors that can be installed inside their own home. To do this, we propose to develop new technologies to enable a better range of motion than existing harness systems, both in the ability to travel between rooms and the ability to support standing and sit-to-stand or stand-to-sit motions.
Estimating Food Intake (EFI)
This project is a joint effort between Drs. Dmitrieva, Behrens, and Winfree. Using an accelerometer, and the NAU High Performance Cluster (HPC), this project seeks to develop the methods of assessing bite-to-bite activities and the contextual behaviors around eating.
The relationship between eating habits and health has been well-documented, demonstrating causal links between poor diet with morbidity and premature mortality. A generation of researchers and practitioners has focused their intervention efforts on the various aspects of nutritional intake (e.g., number and type of calories), yet the process of changing eating behaviors continues to be a challenging task. Studies examining the dynamics and patterns of behaviors associated with eating (e.g., speed of eating movements during a meal, duration of a meal, skipping a meal, and the time of day when a meal occurs) have shown that these behavioral aspects of eating influence health outcomes in an important way. However, these behaviors are typically assessed using self-report methodology, which is prone to underreporting of caloric intake due to social desirability bias, and measurement error due to forgetting the type and amount of foods one has consumed. Approaches to assessment that omit self-report include direct or laboratory observations and biological analyses, but these present their own limitations in that they are often costly, burdensome to participants, difficult to analyze, and require specialized training and/or equipment. The current study proposes to use wrist-worn accelerometers, a technology that has revolutionized the field of physical activity, to develop a novel method for continuous real-time assessment of eating behaviors. First, this project seeks to train a set of machine learning statistical classifiers to reliably identify when someone is eating. Although the motions of the hand during eating are complex, preliminary data presented in this proposal show that multi-axis inertial movement units housed within commodity-based accelerometers have the necessary sensitivity to accurately track these intricate movements. After establishing a set of classifiers that reliably identify eating behaviors, the algorithms will be applied to identify eating behaviors within a new geospatial and social context. At its conclusion, this project will deliver a reliable and valid real-time approach to measuring eating behaviors that is low in cost, user-friendly to researchers, and ecologically valid and low in participant burden and invasiveness.
This line of work will be paradigm-shifting in advancing the scientific understanding of biobehavioral mechanisms of eating behaviors, and may inform the development of targeted real-time interventions that improve public health.
With an ever increasing number of wearable devices in use and an ever improving temporal resolution of the health measures many of these sample, wearable devices are quickly becoming positioned to answer questions not previously possible to even ask. The quantity and quality of these data is expected to yield insights into aspects of daily life that were previously un-observable. This project seeks to develop the methods of analyzing such large quantities of data.
To be able to compare measures from commercially available commodity hardware to that of research grade, one must be able to characterize the relation between the two. This project aims to model a validated physical activity level using inputs of observations from a readily available step monitor.
- Oxidative Stress with Drs. Tinna Truastadottir and Kyle Winfree
- Feasability and Validity of Commodity Heart Rate Monitors with Drs. Traustadottir, Winfree, and Pamela Bosch
- Exercise and the Gut Microbiome with Drs. Greg Caporaso, Deirdre Bycura, and Kyle Winfree
Go Baby Go!
There are many aspects to GoBabyGo at NAU. This includes club activities, research using the cars, and research into harness support systems.
There are currently three active projects related to GBG.
- First, students are developing a data logger able to record when, and for how long, a child uses the car. This project will support an upcoming study investigating the impact of GoBabyGo car use.
- Second, through the support from a student NASA Space Grant and lab funds, students will be developing a harness support system to help enable children with movement disabilities to explore their world.
- Third, through a recently awarded grant proposal, we are developing systems to support stroke survivors (adults) in a house scale harness system.
A study investigating the relation of GoBabyGo car use to changes in functional and cognitive assessments is forth coming.
The NAU GoBabyGo Club (found on Facebook) works to provide modified toy ride on cars to children in the Flagstaff area who have mobility disabilities.
Less Active Projects…
The KineTrax project seeks to develop a highly adaptable robotic data logger, with sufficient sample rate to do full kinematic analysis of human gait from the community setting and the ability to provide real-time feedback and actuation to the wearer. This project integrates a six-axis inertial measurement unit, a wireless network, both analog and digital input, analog and digital output, and an I2C network to sample for a variety of different sources and from several KineTrax devices in a time synchronized manner. Ultimately, the KineTrax device is a research platform for several other projects. This includes a wearable system for gait analysis from a sock, a closed loop system to improve the gait of those with Parkinson’s disease, and even animal tracking with social network analysis.
Related projects include shoe and sock based gait analysis, real-time interventions for Parkinson’s disease, and animal tracking.
KineTrax in still in the development phase and as such does not yet have any formal studies using this device.
Modeling Muscle Structure and Activation (WFM)
This project is a joint effort between Drs. Kiisa Nishikawa and Kyle Winfree. Here, students are designing and fabricating a desktop model of the Winding Filament Model elements, which will work together to physically emulate that which is described by the Winding Filament Hypothesis.
This project is directly related to the prosthetics work of Dr. Nishikawa’s lab.
Current studies include walking while using an active prosthetic device running the WFM instead of the stock controllers.