This is the first edition of a new newsletter about the VA RR&D Accelerometry Project. Our goal is to provide a method of disseminating information to all of our collaborators. Each collaborator, while working independently can enjoy the benefits of sharing research results, and experimental and data analysis methods. This newsletter will provide a forum for discussion, as well as providing news about the status of the project and changes in equipment. This is the web site version of the newletter. You can link it to your own web page or refer to when your paper copy is buried under piles of data on your desktop.
The 1997 RESNA will be held in Pittsburgh from June 21-23. We will present a review of quantitative balance and mobility assesment at the Gerontology Special Interest Group (SIG-16).
"The effects of Tai Chi and strength training on strength balence and sway in elderly frail women" is a new research study looking at the effectiveness of Tai Chi in improving, among other things, balance and sway. The project is part of a new multi center national study funded by the Office of Complementary and Alternative Medicine at Stanford (CAMPS). Wayne Phillips is chairperson of one of the the review groups set up by the Center for Injury Prevention at Stanford University Hospital to look for appropriate topics and questions to investigate relating to what they are calling 'successful aging'. They plan to use the accelerometer as one of their outcome measures. For more information contact: wayne_phillips@scrdp.stanford.edu .
Geoffrey Bush, PhD of the NASA Ames Research Center Vestibular Research Facility will be using the accelerometry system in "Contribution of Proprioceptive and Visual Cues for Maintaining Balance during Stereotyped Movements". He will be studying normal healthy individuals of any age compared to individuals of similar age groups with impaired vestibular function. Malcom Lawton, MD, head of the Head Injury Falls Prevention Program at the Santa Clara Valley Medical Center, San Jose, CA will collaborate. The tests will control monocular, binocular, and occluded viewing distance, and proprioceptive inputs by vibration of various tendons. The activities to be studied are: head shaking, one-foot stand, bend at waist, sit-to-stand, and walking. For more information contact: gbush@mail.arc.nasa.gov
We are very pleased to announce that our proposal 'Reliability and Validity of Accelerometric Gait and Balance Diagnosis' has been approved by VA Merit Review for 3 years starting on October or sooner. In this proposal we will formally establish statistical reliability and validity of a wearable gait and balance analysis instrument by comparison of accelerometry with conventional laboratory gait and balance measures (force platform, goniometry, and video imaging) using subjects having well-defined gender, anthropometric, age, and mobility status.
Accelerometric Analysis of Sit-Stand in Balance - Impaired Elderly SubjectsSubjects - One subject from each group was chosen to be studied for this article; healthy elderly patients, Parkinson's disease patients, pre-hip arthroplasty patients, and stroke patients Equipment Placement - Four 3-axis accelerometers were placed two at the head on eyeglass frames and two at the waist. Testing - Subjects were asked to rise from a chair with arms folded across their chest. Data Analysis - Each 3-axis accelerometer site yielded X, Y, and Z signals. After calibration for individual sensor characteristics and conversion to units of gravity (1 g=9.8 m/sec^2), two calculations were performed:
Results
DiscussionRise time, relative height, and duration of positive and negative peaks are comparable to results found in studies which used force plates and image analysis. The accelerometry-derived trace shows fine detail of head motion (e.g.: 0.5 to 1.0 sec and 1.9 to 2.1 sec) that is not available in forceplate data. The accelerometry system was used to quantify phase-by-phase kinetics and kinematics of sit-to-stand . The acceleration magnitude can be integrated after subtracting the constant 1-g gravitational field. Velocities and displacements of the head were thus calculated for an able-bodied young subject, an elderly hip patient, and a stroke patient, all female. The young subject (A) achieved a peak velocity more than 90 cm/sec while her head traveled about 66 cm; most of the movement occurred within one second. Direction of motion cannot be determined from the acceleration magnitude, but is predominantly vertical and forward. Tilt of the head was assumed to be linear beginning at 0.8 seconds; both the uncompensated and tilt-corrected velocity traces are shown. The stroke patient (B) had a larger difference in head tilt, requiring subtraction of an apparent acceleration of 20 cm/sec^2 (dotted line). With this correction, maximum acceleration was 60 cm/sec over 1.3 sec, for a total displacement of 33 cm. The pre-surgery hip arthroplasty patient (C) had an apparent acceleration of 42.5 cm/sec^2. She used her hands to lift herself from the chair, resulting in an early velocity peak (0.8 - 1.7 sec from start). Peak velocity during extension was 28 cm/sec from 1.7 to 2.8 sec, yielding total displacement of 30 cm. The assumption of linear head tilt is only a partial correction for acceleration so that velocity and displacement can be calculated without artifactual "head motion" after the subject has reached vertical stance. For the first 0.5 - 0.8 seconds, the corrected velocity traces in figures B and C are nearly the same as the value being subtracted, indicating that the correction should not be applied until later. The time at which head tilt occurs can be inferred from examination of other derived values such as DAZX or DAZY VA RR&D Accelerometry Project Current Staff:
|
The vector
magnitude of the right head 3-axis sensor was plotted versus time for four
subjects, each from a different subject group. Curve amplitudes and durations
were compared between the normal versus the 3 impaired subjects.
