0
Research Papers

# A Haptic Simulator for Training the Application of Range of Motion Exercise to Premature Infants

[+] Author and Article Information

Department of Biomedical Engineering, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975kareem.adnan@gmail.com

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440iahmad@uci.edu

Maria Coussens

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440macousse@uci.edu

Alon Eliakim

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440aeliaki2@uci.edu

Susan Gallitto

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440sagallit@uci.edu

Donna Grochow

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440dmgrocho@uci.edu

Robin Koeppel

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440rkoeppel@uci.edu

Dan Nemet

Child Health and Sports Center, Pediatrics Meir Medical Center, 44281, Israeldnemet@uci.edu

Julia Rich

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440jkrich@uci.edu

Feizal Waffarn

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440fwaffarn@uci.edu

Dan M. Cooper

Department of Pediatrics, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975 91440dcooper@uci.edu

David J. Reinkensmeyer

Department of Biomedical Engineering, Department of Mechanical and Aerospace Engineering, University of California at Irvine, 4200 Engineering Gateway, Irvine, CA 92697-3975;dreinken@uci.edu

J. Med. Devices 3(4), 041008 (Dec 09, 2009) (7 pages) doi:10.1115/1.4000430 History: Received September 02, 2008; Revised October 06, 2009; Published December 09, 2009; Online December 09, 2009

## Abstract

The range of motion exercise is an experimental therapy for improving bone and muscle growth in premature infants but little is known about the magnitude of pressures that must be applied to the limbs during this exercise to elicit a physiological benefit and novice caregivers currently must rely on subjective instruction to learn to apply appropriate pressures. The goal of this study was to quantify the pressures applied by experienced caregivers during application of this exercise and to create a haptic simulator that could be used to train novice caregivers such as parents to apply the same pressures. We quantified the pressure applied by two neonatal intensive care nurses (“experts”) to the wrists of nine newborn, premature infants of varying gestational ages using an infant blood pressure cuff modified to act as a finger pressure sensor. The experts applied statistically significant different pressures depending on gestational age but did not differ significantly between themselves in the pressure they applied. We then created a robotic simulator of the premature infant wrist and programmed it to respond with the measured pressure-angle properties of the actual infants’ wrists. The novice adult participants $(n=19)$ used the simulator to learn to apply target pressures for simulated wrists that corresponded to three different gestational ages. Training with the simulator for 30 min allowed the participants to learn to apply pressures significantly more like those of the experts. The performance improvement persisted at a retention test several days later. These results quantify for the first time the pressures applied during assisted exercise, include novel observations about joint flexibility and maturation early in life and suggest a strategy for teaching exercise intervention teams to provide assisted exercise within a more reproducible range using haptic simulation technology.

<>

## Figures

Figure 1

Left: Infant blood pressure cuffs worn on the thumb and index finger of the clinician. Right: Experiment set-up—the nurse ranged the infant’s wrist as we recorded pressure readings and wrist movement with a video camera.

Figure 2

Example pressure-wrist angle curve for one infant. Individual data points from each video capture frame are represented as dots, for five repetitions of the exercise. Polynomial curve fit is shown by the thick line.

Figure 3

Robotic baby wrist simulator. From left to right: 3D CAD model of mechanical components of robot, CAD model of assembled components, picture of actual robot showing silicone covering hand and arm to create a more realistic feel.

Figure 4

Feedback screen for the haptic training program. Images of the babies were shown to participants to indicate which baby the robot was behaving like. During the training phase participants were given knowledge of results in the form of a light indicating if the previous trial was: too soft, too hard, or good with a yellow, red, or green circle shown, respectively. The large number is a counter that indicated which trial number the participant had just completed.

Figure 5

Pressure-wrist angle curves. The first three graphs show each group’s pressure versus angle curves for each nurse, infant, and wrist within that group, where group 1 has the youngest infants (gestational ages for group 1: 30–33 weeks, group 2: 34–37 weeks, and group 3: 38–41 weeks). The dashed line is the mean curve for all the curves in that group. The fourth graph shows only the mean curves.

Figure 6

Average maximum (flexion) and minimum (extension) stretching pressure applied to the infant wrist, as a function of infant group (i.e., infant age). Nurse 1 is denoted with an x and nurse 2 with an o. Pressure depended significantly on infant age (ANOVA, linear contrast, p<0.02). The bars show 1 standard deviation across infants (three in each group) and wrists (left, right).

Figure 7

Left: mean pressure matching error. The mean pressure error across participants during the baseline, short-term retention test, and long-term retention test with bars used to denote the standard deviation of the force error across participants. Right: mean of the absolute pressure matching error. Averages of absolute value of error and with error bars to indicate one standard deviation. Participants showed significant improvement as a result of training.

Figure 8

Variability of pressure matching error. Standard deviation of error averaged across all participants and all babies. Error bars indicate one standard deviation of the individual standard deviations. Significant reduction (p=0.01) of the average standard deviation from baseline to retention test indicates that participants became more consistent at applying the desired pressure with training.

Figure 9

Actual pressure versus target pressure. These plots show each participant’s average pressure applied to each infant (open circles) and the mean of all participants’ average pressure for each infant (filled circles near regression lines) versus the target pressure for each infant.

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Topic Collections