An Engineering-Based Balance Assessment and Training Platform

Overview

This is a proposal to develop a Balanced Reach Training Protocol (BRTP) to evaluate and train dynamic standing balance. The BRTP is based upon the Balanced Reach Test (BRT) that the investigators previously developed and validated. In the BRT subjects stand and point to a target disk moving unpredictably across a large projection screen in front of them without stepping. Body movements undertaken to track the disk are integral to many daily activities and represent an important class of "expected" balance disturbances that can precipitate falls. The BRTP employs engineering and psychophysical methods, and exploits advances in real time computing in a novel and innovative way to more effectively evaluate and train balance function. The BRTP presents a challenging reaching/tracking task that subjects perform at their limit of balance. The BRTP is an objective, quantitative test that can evaluate balance function without floor or ceiling effects, and train balance across the spectrum of aging, disease, and injury.

This study comprises two Specific Aims: SA-1) Development of a Balance Assessment and Training Platform (BATP) based upon the existing BRT; and SA-2) Evaluation of the BATP Assessment Module's test-retest reliability and the Training Module's motor learning effect. Methods related to Specific Aim 1 SA-1.1: Develop a BATP based on the BRT that incorporates real-time computing and an Adaptive Staircase Algorithm to establish subjects' Limit of Balance. Modify the BRT's experimental control algorithm and hardware setup such that data from our extant measurement systems are continuously streamed to a Real Time Computing Workstation and read by the experimental control algorithm, enabling it to monitor foot movement and compute key performance measures on an ongoing basis. These measures will include Root Mean Square tracking Error (RMSE) and Root Mean Squared Deviation (RMSD) between the ground plane projection of Center of Mass (CoM) and the center of the Base of Support (BoS). Incorporate an Adaptive Staircase Algorithm into the real time experimental control algorithm to establish the target positions corresponding to Limit of Balance, based on foot movement (i.e., stepping). SA-1.2: Replace the projection screen with Virtual Reality eyewear. The experimental control algorithm will be further modified to display a virtual sphere in the virtual reality eyewear. The virtual reality eyewear will provide a direct view of the actual physical environment but also show the sphere following the trajectory specified by the Adaptive Staircase Algorithm, or the assessment or training modules. When the subject's tracking fingertip contacts the center of the target sphere it will provide visual feedback by changing color. SA-1.3: Develop the BATP's Assessment and Training modules. The Assessment Module consists of measuring the subject's Limit of Balance, computing the manifold bounding target motion, and the target's motion within the manifold; as just described. The harnessed subject then tracks the disk for 90 sec. After a 2 1/2 minute rest the process is repeated a second time. Tracking and resting durations can be varied to optimize measurement quality and accommodate individual subject needs and capabilities. The Training Module operates in a loop that presents an operator-specified number of training bouts. One training bout consists of measuring Limit of Balance, computing the manifold and target motion, presenting the tracking target for five minutes, and pausing for a 2 1/2 minute rest period before signaling the operator to begin the next bout. Tracking and resting durations can be varied to optimize training efficacy and accommodate individual subject needs and capabilities. For this proposal a training session will consist of five bouts and last approximately 60 minutes. As in the Assessment Module, subjects will be harnessed while performing the tracking task to guard against falling. The Assessment and Training modules both collect the same data (All data will be saved for offline, post-test analyses: a) The values of r and corresponding to each Limit of Balance, and Limit of Balance; b) The motion (position vs. time) of the target sphere; c) The motion of each of 13 body segments, as well as the tip of the tracking finger \[these data will be used to compute tracking error (RMSE), balance stability (RMSD), and to inform the Adaptive Staircase Algorithm when stepping occurs\]; d) Ground reaction forces and moments for each foot. Methods related to Specific Aim 2 SA-2.1: Evaluate the Assessment Module's test-retest reliability and compare subject performance in it to that in established clinical measures of balance, reach, and Fear of Falling. Twenty stroke (STR) adults will undergo two tests with the Assessment Module 48 hours apart. Clinical measures will be administered before each test. Test-retest reliability will be evaluated SA-2.2: Evaluate the Training Module's motor learning effect of three one-hour training sessions over one week. Ten subjects randomly selected from those that participated in SA-2.1 will undergo three one-hour training sessions for one week. The SA-2.1 test-retest data will also serve here as double baseline data to establish before-training balance performance. Subjects will be evaluated using the clinical measures at the beginning and immediately after training. Motor learning will be assessed using data from the baseline assessments, the clinical measures, and from the Training Module taken from the last 90 sec of the last training bout of each training session.