by James Carollo, Ph.D., P.E. Director, Center for Gait and Movement Analysis

The Children’s Hospital, Denver

Assistant Professor, Rehabilitation Medicine and Orthopedic Surgery Departments

University of Colorado Health Sciences Center

Through the efforts of co-medical directors Dr. Dennis Matthews, Chairman of Rehabilitation Medicine, and Dr. Frank Chang, Director of Orthopedics, a consortium was assembled that included members of the hospital’s orthopedic surgery and rehabilitation medicine departments, physical therapy faculty members from the University of Colorado Health Sciences Center (UCHSC), and engineers from the University of Colorado at Denver and Colorado School of Mines.  This, along with a mix of funding from Children’s Hospital, UCHSC, and several private foundations, permitted the project to go forward.  Design of the facility began in late 1998, with construction taking place between March and June of 1999.   The demand for the facility required that clinical operations be initiated as quickly as possible, so the normal installation and testing portion common to all new facilities was compressed into a six-week timeframe to meet the July 1 opening requirement. While this was a challenge for all members of the team (see Figure 1), planning, past experience, and the reliability of modern gait analysis instrumentation made this possible with only minor setbacks, none of which was sufficient to delay the start of clinical operations.

One of the great benefits of building a new facility where no facility has existed previously is that you can start with a clean slate and design the facility without needing to incorporate prior methodologies or inherited instrumentation. Our goal was to provide maximum flexibility to meet the varied research interests of the consortium, while still providing high quality measurements to support the most immediate need: quantitative clinical gait analysis. 

Figure 2: View of the laboratory from within the control room, showing observational video system on the right, Vicon 512 datastation on the left, and data collection area in the background.

The laboratory is located in a 2000 square foot area in the lowest level of Children’s Hospital (see Figure 2).  Since this area had been used only for storage prior to construction, we were able to identify favorable kinematic and observational video camera locations within the total unfinished volume, and then build dividing walls as required to break up the space into appropriate work areas. A special urethane rubber floor tile was used to eliminate marker reflections during motion capture, and delineate the primary gait lane down the center of the laboratory (see Figure 3).  Motion data are captured with a 6 camera Vicon 512 system, using progressive scan 240Hz cameras.  For most routine gait analyses, these cameras are set for 120Hz so that maximum vertical resolution can be preserved. The cameras are arrayed around the primary gait lane, using three on each side and avoiding camera rays that directly align with the direction of forward progression. For routine studies, oblique cameras located in each corner are mounted to studio lighting clamps that permit vertical height adjustment along a fixed pipe mounted to the wall.  They can be easily removed and set on tripods for experiments that require a smaller measurement volume.

Figure 3: Physical therapist Bobbie Riedner prepares a subject for testing.

Since gait and movement analysis at the hospital prior to laboratory construction had been based predominantly on visual assessment, it was important to include a somewhat more elaborate observational video system in the new facility. Observational videos are recorded initially during our screening clinic, without markers or electrodes, and using a separate gait lane that is closer to the laboratory’s back wall and illuminated more appropriately for color video recordings. The system uses two Panasonic AW-E300 digital, 3 chip color cameras each mounted to an integrated pan & tilt controller, so that the lateral and fore/aft views recorded in the measurement area can be controlled by personnel who are not in the test subject’s line of sight. This minimizes possible distraction during a session, and helps our physical therapist elicit a more characteristic gait pattern from young subjects. An Echolabs special effects generator is used to provide a split screen effect between the 2 cameras, and a Videonics graphics generator permits character annotation and event markers to be introduced over the video for identification. The video output is recorded using Panasonic DV2000 digital VCRs, and during kinematic recording, is routed to a video digitizing board in the Vicon workstation to enable synchronized movie and motion capture.  Since the video is recorded using a digital tape format, important segments can be transferred without image degradation via Firewire to a non-linear, computer-based editing system to create pre-op/post-op summary records and movie archives.  Our eventual plan is to store full patient videos online and accessible via our laboratory database and campus network.

Figure 4: Force platform array showing 1:2:1 configuration, mounting plate used to accomodate different configurations, and suspended floor grid.

Ground reaction forces are recorded using four Kistler 9281CA piezoelectric force platforms, arranged in a somewhat non-traditional arrangement (see Figure 4).  To increase the likelihood of obtaining both a left and right foot-strike in a single pass, the force platforms are oriented with the first platform’s long axis perpendicular to the gait lane, the next two rotated 90¡ and mounted side-by-side, and the last platform aligned with the first.  This produces a 1:2:1 force platform array that in practice seems to work equally well for both children and adults.  Nevertheless, if an experiment requires a different configuration, the platforms can be easily unbolted and arranged in at least 20 different patterns using predrilled holes on the 8 foot custom mounting plate. To help prevent targeting, each platform is covered with the same rubber flooring material that identifies the central gait lane, so that the array blends into the walkway.  While covering the platforms reduces their natural frequency, impact testing at Kistler showed that there was less than a 20% reduction with this relatively stiff flooring material.

We selected Motion Lab System’s MA-300 for our dynamic EMG recordings, primarily due to the historical reliability of earlier generations of the system.  Our only modification to the device is that we use neonatal gel-type surface electrodes rather than the standard dry-type pads, so that we can employ a smaller active area and inter-electrode distance than that provided in the standard system.  This seems to reduce crosstalk from surrounding muscles, and helps us limit the number of fine-wire electrodes required during routine procedures. 

Figure 5: Synchronized Video/EMG display. EMG traces at the far right of the display correspond to the instant captured by the video. The EMGs then scroll right to left in successive frames. Notice that this "stiff knee" gait patient shows inappropriate rectus femoris activity in mid-swing (top trace).

In addition to the standard EMG analog recording provided by Vicon, we have developed a custom data acquisition system that provides real-time display of all EMG channels superimposed directly over the observational video record of the subject (see Figure 5).   This has proven especially valuable for trying to understand complex motor control issues associated with CP, stroke, and TBI, and helps us monitor surface electrode isolation during the recorded EMG muscle test.  We have even found that subjects who have difficulty producing isolated muscle contractions on their own can use the feedback provided by the synchronized video/EMG to reduce co-contraction at least temporarily.  From an analysis standpoint, adding video/ EMG to Vicon’s Polygon movie pane during a motion capture trial provides a remarkably useful tool for integrating all our measurements into a single application (see Figure 6).  It is our feeling that this improves our understanding of the laboratory results, and reduces the amount of analysis time required for each patient.

Figure 6: A typical Polygon screen used at CGMA during analysis, with all measurements available in a single application.

Clinical gait analysis in Colorado has taken an important step forward in the nine months since opening this facility, and recently we celebrated an important milestone with our 100th clinical patient (see Figure 6).  Erica is a really special kid, who has had a long history with The Children’s Hospital in addition to being an active participant in our Handicapped Ski Program at Winter Park.  She is now one of a growing number of kids who have both on-mountain skiing videos and a full gait analysis.  It is for adults and children like Erica, who have high expectations for an active lifestyle, that we have invested in the tools necessary to better understand their overall patterns of movement. With these tools and the expertise of the team, we feel confident that we can now help each of them reach their full potential.