In Motion

Running Analysis - Different Technologies - Different Processes - Different Results

Posted by Oliver Watson

May 3, 2017 10:34:20 AM

Choices, choices:

In this post we will go over some of the various technologies available to clinicians, providers, and coaches for measuring and analyzing runners' gait. Offering running analysis as a cash business is extremely lucrative for clinicians with large running populations, and while no single solution is perfect for everybody, we hope that this post will serve as a means to guide professionals towards the right technology that fits into their processes, yields the results they require, and helps them perform at their highest level possible.

Video Analysis


Video Analysis technology has been a common fixture in clinics offering running analysis for many years. Technology ranges from highly sophisticated, multi-camera setups, to iPad video capture and analysis. Often times, the key to running an effective video analysis is following a standardized process - it is one thing to just film a runner on the treadmill willy-nilly and make some determinations on his or her gait and mechanics, but in our experience the most effective users of video analysis have a specific plan for every runner, follow their protocol without fail, and ensure that they are maintaining consistent capture perspectives - this last one is extremely important when comparing data captured at different times.


  • Rich, vivid data: everybody likes to see themselves on video (for the most part), and this can make it easier to explain to a runner what you're interpreting from their data, especially if you can show them a video of what's going on as evidence for your assessment.
  • Flexible setup: since we are analyzing in 2D, users can often get away with as few as one or two cameras - in most cases we see clinicians analyzing runners from the Posterior and Sagittal perspectives - which means that you don't necessarily need to have a large designated space for your video captures.
  • Online analysis: over the past few years we have seen a lot of clinicians and running coaches offering "Online Video Analysis;" i.e. the runner follows a strict procedure for setting up his or her phone/camera in order to collect video data without having to go into a clinic. The video is then uploaded to the provider who subsequently conducts the analysis. This can be a great service for runners who want to improve their performance, but are unable to make time for an in-person session. It is important to note that clear instructions must be provided in order to ensure that usable data is collected.



  • Rich, vivid data: although this is also one of the pros, video data - especially high-speed, high resolution trials - can be a lot to handle. Analyzing the data can take hours, depending on the clinician, and can lead to the dreaded call to customer support with the phrase: "my computer's hard drive is full, what do I do?" (Hint: crop your videos down to cut out unnecessary video data, and archive older trials on backup or network access drives.)
  • Accuracy: In order to achieve a usable degree of accuracy, very specific steps must be taken in order to ensure that your camera perspectives are perpendicular to the center of the runner's plane of motion. For example, if you measure a runner's knee angle, but filmed the video halfway between the Posterior and Sagittal perspectives, you might as well toss that measurement in the trash, because it will NOT be accurate in any regard. Video captured from any perspective can be useful, analysts just need to make sure they keep in mind when graphical measurements will be accurate (sometimes, under the right circumstances and with the correct controls in place), and when they won't.
  • Human Error: One problem that can come up during running evaluations with video relates to misinterpretations of the data, or overreaching. If you can't back up an interpretation with an accurate, sound measurement or observation, then don't say it definitively. If you want to be able to interpret every single minute detail about a runner's stride, then we have some advice for you - invest in some form of 3D analysis!


3D Motion Capture - Marker Based


Once limited to university research labs, 3D Motion Capture technology has recently become much more accessible for clinicians and coaches alike. Through the use of retro-reflective markers placed on various joints on the runner's body and multiple (at least 2, but typically 4 to 8, calibrated infrared cameras, 3D Motion Capture systems are able to triangulate the positions of each marker in space as a subject runs and create a 3D skeletal model of their body. This data can provide an extremely granular look at how a runner moves to a skilled analyst.


  • Accuracy: 3D motion capture data is about as accurate as it gets - assuming all of the joint markers have been placed correctly and that an appropriate marker set is selected for the analysis. A running analyst can obtain a snapshot of exactly how a person is moving and have the ability to analyze and manipulate all of the data.
  • Automation: Since we are talking about positional data, most motion capture software suites (if not all) are able to instantaneously process the data, and in many cases are able to build out highly detailed, user-defined reports, all with the click of a button.  This cuts down on the amount of "busy" work required by the analyst, such as formatting report templates and ensuring that pages print out correctly.


  • Setup: One of the major drawbacks that comes with 3D Motion Capture is its required setup time. Depending on the marker set used for the analysis, a clinician can be required to put as many as 60 markers onto the runner's body - and if one of these markers is incorrectly placed by a few centimeters, it can throw off the entire analysis and render the collected data useless.
  • Fixed Configuration: in most cases, in order to run a full-body 3D capture of a runner, cameras must be mounted in permanent locations around the capture space. This can be costly and time consuming, but does have the added advantage of keeping the cameras out of the way - i.e. it is very, very hard to kick over and destroy a camera that is mounted 8 feet up on the wall, believe me, I've tried.
  • Indoors only: With the exception of some higher end motion capture systems, users are required to run analyses indoors (usually on a treadmill), as sunlight can have adverse effects on the infrared emitting and receiving aspects of motion capture cameras. If you want to run a 3D motion capture outdoors, we suggest going with IMUs (Inertial Measurement Units) or be prepared to shell out $75k-$200k for a high end motion capture system, which might not be practical for most clinicians.

Sensor Based - IMUs


Inertial Measurement Unit (IMU) based motion capture is very similar to marker-based motion capture with one fundamental difference - instead of placing markers on the joints, tri-axial (meaning that they contain sensors that measure in all three planes of motion, x, y, and z) IMUs are placed on the various segments of the body. IMUs are pretty ubiquitous in our daily lives. Have you ever played a racing game on your smartphone in which turning the phone controls how the car turns? Thank the underlying technology behind IMUs! Over the last decade, advances in terms of sensor size and wireless transmission have drastically improved IMU performance with regards to accuracy, reliability, and portability. With IMUs, you are able to collect data that is strikingly similar to 3D marker-based motion capture, in the lab or in the field.


  • Flexibility: Conduct analyses practically anywhere - in the lab or out in the field. Data capture can be conducted in normal conditions for the runner - whether on a track, asphalt, or a trail setting.
  • Automatic, Custom Reports: Similar to marker-based Mocap, you can analyze the data and build out reports using built-in system algorithms. Accompanying software can also automatically identify events during the gait cycle, which can reduce time spent manually manipulating data and allow analysts to focus on interpreting the data and coming to conclusions in order to improve the runner's mechanics.


  • Intrusiveness: since the athlete has sensors strapped to his or her body segments, one could infer that this could interfere with natural body movement. This can cause anomalies because the main goal of running analysis is to dig into what is wrong with a runner's normal gait, but if the runner is altering his or her gait due to an external factor, it might have a negative effect on the end results of the analysis.
  • Magnetometers and treadmills don't always get along: IMUs typically consist of three main components: an Accelerometer, Gyroscope, and a Magnetometer. Magnetometers are used to determine where each sensor is in relation to the other sensors, but when you throw a treadmill motor into the mix you can get some interesting abnormalities in your collected data.

Markerless Motion Capture


Markerless Motion Capture is able to collect a wealth of data on runners in controlled conditions without the need for manual interpretation or marker placement. Marker-less motion capture systems typically use one of two different fundamental technologies: depth sensing cameras or advanced edge detection technology. Depth sensing cameras emit infrared beams which reflect off of the runner and are picked up by the camera's IR sensor - at different distances the reflected beams display as different colors, which are interpreted (through the use of advanced mathematical algorithms) by the software to form a skeletal model. Edge Detection, on the other hand, can be described in simple terms as a patchwork of video views that are analyzed and combined in order to create a model of the subject. Basically, if a subject is standing in the center of a marker-less motion capture system's capture volume that uses edge detection, each camera perspective (typically 8 cameras or more) is able to determine where the edges of the subject's body are in space, and advanced software stitches all of the data from each perspective together in order to create a full 3D view. Since Edge Detection technology is more commonly used for advanced research and animation, and less so in running analysis, we will focus on markerless motion capture using depth sensing cameras for the following pros and cons.


  • Automation: In our experience, marker-less systems are able to collect, analyze, and report on running performance without the need for excessive user input.
  • Advanced data sets: As the systems often operate automatically, they can be programmed with inverse dynamics calculations that allow analysts to view data on runner's ground reaction forces and joint loading parameters without the need for costly instrumented force/pressure treadmills.
  • Speed: Since there are often minimal setup criteria (outside of system calibration), such as placing markers, clinicians are able to obtain meaningful, accurate data on running performance in mere minutes.
  • Repeatable Testing: When the results of a capture are generated in minutes, it is fair to say that you can easily run subjects through multiple tests during a session, and compare the results in near real-time. This allows for clinicians and providers to spend less time measuring angles and making sense of kinematic data, which in turn enables them to interface with the runner and educate them on how to improve their mechanics, and then turn around and test again to see if the suggestion or intervention yielded the desired results.


  • Fixed installation: In order for depth sensing cameras to be setup and calibrated correctly, they usually need to installed at fixed distances from where the subject will be running. Changing the location of the cameras can change the output, which is something we want to avoid (consistency in analysis is key), especially when repeating tests during follow up appointments with runners.
  • Strict Clothing Requirements: Marker-less systems track each of the body's segments as a whole, so if a runner is wearing baggy shorts or pants during an analysis (we're not sure why someone having a running analysis would wear baggy pants, but it definitely has happened) the system could interpret the sway of the loose clothing as additional movement of the limb - resulting in lower accuracy results.

As we mentioned before, offering running analysis services to your existing patients or athletes can be a great way to boost revenue and spread the word about your business. Adding technology to the mix only enhances the experience for runners and will keep them coming back, but only if implemented properly.

If you are interested in adding a running analysis program to your clinic, running club, or performance center and have any questions, get in touch with us by clicking the  "Contact Us Today" button below. We can offer suggestions specific to your vision as well as provide information about the solutions we offer.

Contact Us Today!

If you have any questions or feedback, we would love to discuss with you!  Feel free to leave a comment below!

Topics: How-To Posts and Guides

Recent Posts

Subscribe to Email Updates

The zFlo blog is maintained by our team in hopes of providing some insight into the motion-analysis industry. We work with a large number of companies and have over 10 years of experience working with video and optical motion capture technology - this is where we do our best to explain some of the more important aspects of how this technology can be used.