Graduation Year


Document Type

Thesis - Campus Access Only

Degree Name

Bachelor of Science – Health and Human Performance

School or Department

Health and Human Performance


Health and Human Performance

Faculty Mentor Department

Health and Human Performance

Faculty Mentor

Dr. Matthew Bundle


Accelerometer, geriatric, bed-fall

Subject Categories

Analytical, Diagnostic and Therapeutic Techniques and Equipment | Medicine and Health Sciences



Rolle, Brielle, B.S., May 2017 Health and Human Performance Exercise Science

Analysis of 3-Dimensional Movement Kinematics of Rising from Supine: Implications for Bed-Fall Detection

Faculty Mentor: Dr. Matthew Bundle

PURPOSE: This is the initial phase of a multi-part study to develop an accelerometer-based wearable device designed to detect and distinguish human movement patterns that precede falls. The goal of the device is to prevent falls by alerting medical personnel prior to a fall. Bed-fall can be attributed to factors including compromised cognitive state, physical limitations, or muscular imbalance which impairs normal biomechanics. 60% of nursing home residents fall each year, and half of these patients experience multiple falling episodes (5). Injury and illness associated with falls take a substantial fiscal toll on our economy. Annual costs associated with falls total $34 billion in the United States (6).

METHODS: Our subject population consisted of healthy young adults ages 18-25. Subjects were marked at eight anatomical landmarks with reflective indicators. They were filmed rising from supine to standing using internally synchronized high-speed motion capture cameras at a rate of 125hz. The frames obtained from the high-speed motion capture system were integrated into the DLTcal5 and DLTdv5 programs in MATLAB designed and described by Tyson Hedrick and modified by Clarke and Weyand. The program provides a precise digitizing environment with numerical output and is designed to acquire three-dimensional coordinates from multiple video sources calibrated via a set of direct linear transformation (DLT) lens coefficients (10). Instantaneous marker velocities were obtained from the scalar length of the vector between the two 3Dcoordinate locations present in sequential video images and the period elapsed between 2 video frames (i.e. 0.008 s).

RESULTS: These findings quantify the movement patterns used during a rise from supine to standing. The data indicate that the shoulder provides consistent displacements between subjects that are of a magnitude sufficient to warrant additional consideration for future instrumentation. The lower leg consistently displayed the greatest velocities with the metatarsal joint having an average peak velocity of 1714 ± 162 mm/sec (Fig 5).

FUTURE DIRECTIONS: The next phase of this study will examine elderly populations to determine whether their movement patterns are similar to those of young adults.

Honors College Research Project




© Copyright 2017 Brielle E. Rolle