Collecting Experimental Data

The sections below outline our suggestions for collecting high quality experimental data for use in analyzing human and animal motion and generating simulations.

When you start collecting experimental data to analyze motion and generate dynamic simulations, we highly recommend that you develop and document lab protocols and standards. By establishing protocol for marker sets, camera locations, and force plate coordinate frames, data will be more repeatable and it will be easier to pre-process and import data into OpenSim.

Take Photos and Video during Data Collection

• Digital cameras and camcorders are affordable! Take lots of photos and video during experiments so that you can verify marker placement and other factors for the data you collect. Get a tripod and record each trial. Some motion capture systems can even sync to a digital camcorder.
• Take pictures of your subjects in the static pose. These picture are valuable for evaluating the results of the Scale tool and creating Marker Sets.

Marker Set for Collection of Full-body Motion Capture Data

This section was written by Chand John, with additional editing by Sam Hamner

If we know the positions of three points on a rigid body in three-dimensional space, we can completely determine that rigid body's position and orientation. Our marker set is designed to give us the locations of at least three markers on each rigid body segment so that the position and orientation of each body segment of interest can be completely determined. This marker set is a modification of the Cleveland Clinic marker set. The marker set also includes two additional markers on each ankle (medial and lateral malleoli) and on each knee (medial and lateral epicondyles), which can help determine the ankle and knee joint centers. Note that these eight additional markers were only used in static trials and range of motion trials, and not in the dynamic gait trials.

BODY SEGMENTS
There are 12 body segments of interest to us:

1. Torso (and head)
2. Right upper arm
3. Right lower arm (and wrist)
4. Left upper arm
5. Left lower arm (and wrist)
6. Pelvis
7. Right thigh
8. Left thigh
9. Right shank
10. Left shank
11. Right foot
12. Left foot

MARKER SET (50 MARKERS)

OTHER TIPS

• • As a general rule, you must place enough markers on your subject to scale and track each body segment you will model. You need at least three non-collinear markers to track the 6 DOF motion (position and orientation) of a body segment.
  • Try to place markers on anatomical locations that will have the least skin and muscle motion.
  • There are several OpenSim model marker sets that you can use and adapt, including:
    
    • Full-body model used for dynamic simulations of running by Hamner et al., 2010: https://simtk.org/project/xml/downloads.xml?group_id=516
  • To read more about markers sets for motion capture, please see the following references:
    
    • Cappozzo, A., Catani, F., Croce, U.D., Leardini, A., 1995. Position and orientation in space of bones during movement: anatomical frame definition and determination. Clinical Biomechanics (Bristol, Avon) 10, 171–178.

    • Davis R, Ounpuu S, Tyburski D, Gage J. A gait analysis data collection and reduction technique. Hum Mov Sci 1991;10:575–87.

    • Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., 1990. Measurement of lower extremity kinematics during level walking. J. Orthop. Res. 8 (3), 383–392.

Measuring Subject Characteristics

• Measure as many subject specifics as possible, including height, mass, body segment lengths, mass distribution (if DXA is available), and strength (if a Biodex is available). 
• You can use this data, along with marker positions, to best match the generic model to a specific subject.

Functional Joint Centers

• To improve scaling of the torso and lower extremities, you can calculate functional joint centers and append the joint centers to your static trial data (see __40__Associating Data with a Motion).  
• Have your subjects perform movements to calculate functional joint centers at the hip, knee, ankle, and/or shoulders.
  
  • Hips & Shoulders: Have the subject move each arm/leg in a star pattern, raising and lowering their arm/leg 4 or 5 times as they rotate from an anterior position to a posterior position.
  • Knee: Have the subject do a knee squat. Repeat 3 times.
  • Ankle: Have the subject plantarflex their ankle and stand on their tip toes, then lower. Repeat 3 times.
• For more information about functional joint centers, please see the following references:
  

  • Gamage, S.S., Lasenby, J., 2002. New least squares solutions for estimating the average centre of rotation and the axis of rotation. J Biomech 35, 87-93. 

  • Siston, R.A., Delp, S.L. 2006. Evaluation of a new algorithm to determine the hip joint center. J Biomech 39, 125-130.

Measuring External Forces

• You must measure all external forces applied to your subject in order to model the full dynamics of the system. This includes ground reaction forces and any forces applied by external objects like a brace or harness. 
• Record or otherwise determine the magnitude, direction, and point of application for all forces applied to the subject. Make sure you also document what coordinate system the forces are measured in so that you can perform any necessary transformations when you prepare your data for import into OpenSim.
• Calibrate center-of-pressure measurements with marker positions using a calibration "T" to pinpoint where on the force-plate the point load (lowest tip of the "T") is being applied. If the center-of-pressure calculated from the force-plate does not match the "T" location from markers (within marker resolution), you need align the force-plate and marker mo-cap reference frames.
• Note that generating dynamic simulations of motion are very sensitive to noise in force plate data whether due to building vibration, treadmill vibration (if using instrumented treadmills), or other sources. The baseline noise levels of force plates should be carefully evaluated before pursuing a simulation-based study and filtering of the data may be necessary to generate simulations.

Measuring EMG

• If possible, you should collect EMG data for as many muscles as possible, particularly the muscles whose function you are most interested in studying.
• Collecting maximum voluntary contractions or other measurements for normalization can be very useful for comparing estimated activations from simulations to EMG data.
• If you don't have access to EMG, search the literature for reported values for similar activities.
• Access to EMG data from experimental motion capture or the literature will allow you to evaluate whether the muscle activations you calculate in a simulation are reasonable.
• For more information about collecting EMG, please see the following references:
  
  • http://www.noraxon.com/emg/emg.php3: A link to the ABC of EMG booklet. This describes many useful ways to collect and filter EMG data. Free download.
  • Hermens, H. J., Freriks, B., Disselhorst-Klug, C. and Rau, G. (2000). Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 10, 361-74.
    
    

Contributors: Samuel Hamner, Jen Hicks, Tim Dorn, Kat Steele, Chand John