In this interview, News Medical speaks with Jason Rogers about using CatWalk, RunwayScan, and Ink & Paper for rodent gait analysis and gait measurements.
Please could you provide an overview of the work that you do regarding rodent gait analysis?
For the past fourteen years, I have worked with and trained researchers on how to use Catwalk XT, a free-walking, automated gait analysis system.
What are the primary ways in which a rodent’s gait can be measured?
The traditional methods are observational scoring (i.e., BBB), running wheels, and ink and paper. Automated methods include free-walking systems, such as CatWalk XT and RunwayScan, or treadmill systems, like DigiGait and TreadScan.
Why measure a rodent’s gait in the first place?
Locomotion is a major behavioral endpoint that has important applications in research. Changes to locomotor behavior, measured by the animal’s gait, can be affected due to age, genetic variation, or from changes within the central nervous system, peripheral nervous system, or from the muscles or bones themselves.
What kind of data can be produced by measuring a rodent’s gait?
One major advantage to automated systems is the volume of data produced. Most commercially available systems, free-walking or treadmill, provide over fifty parameters, such as stand time, swing time, swing speed, base of support, print length, width, area, inter-limb coordination (phase dispersion), regularity index, and cadence, for example. A rodent’s gait is a complex process and it is important to offer many parameters to accurately describe the locomotor patterns displayed.
What kinds of systems do you have on offer to measure a rodent’s gait?
Our complete, turnkey system is called CatWalk XT. It uses Illuminated Footprints technology based on the voluntary movement of the animals. The system has a high-speed camera (100 fps), and our software automatically classifies and quantifies the animal’s gait. CatWalk XT has over 50 built-in parameters with two- and three-dimensional visualizations. The software is a ‘white box,’ meaning that you have access to the raw data. Finally, we offer a goal box for appetitive and naturalistic motivation.
Are there any specific drawbacks to TreadScan?
I have never seen or used TreadScan, so it is important to note that what I mention is based on published data from researchers in the field.
That said, the experience in the literature reports that it can be quite difficult to master, meaning novice users have a more difficult user experience. It is also more difficult to maintain uniform lighting and to keep clean. Finally, it was reported that about 45% of the data could be reproduced. Again, I would check the literature and contact the authors of those studies for their personal experiences. From what I have seen, there are some benefits, such as the visualizations and data parameters, such as stride length, paw pressure, spacing, running speed and so on.
Image Credit: Noldus Information Technology
Could you please discuss the basic premise of the automated systems and how they work?
An automated system simply means that some sort of technology is analyzing the rodent’s gait and generating the data, as opposed to a human observer. I think the question refers more to the basic premise of a free-walking system, which I am happy to discuss.
The basic premise of a free-walking system, like CatWalk XT, is that the animal walks along a green LED-illuminated walkway. Overhead, a ceiling provides red light, creating an outline of the animal as it walks. Underneath, a high speed camera is mounted to the frame. As the animal walks along the walkway, the paws reflect light towards the camera. Finally, the system has a control unit which is attached to a computer.
Are there any drawbacks to this kind of free walking system?
One major assumption in gait analysis is that there is a gait to analyze. Animals incapable of walking will not produce any data, regardless of the system used to analyze it.
Are there any other drawbacks of free walking? Do these systems struggle if the animal drags a paw, for instance, rather than walking?
As long as the animal makes contact with the walkway, you are going to generate data. Dragging a paw means you have nothing but stand data (no swing), but you will have data nonetheless. In addition, you will have data for the other three paws compensating for the damaged limb.
One drawback manufactured by our competitors is in the speed of the animal. They claim that maintaining consistent speed is the primary advantage of a treadmill; however, that is not the case with all systems. Nonetheless, there are experts on both sides of the ‘normalize speed’ fence, as it were, and the literature has many opinions on this matter.
How can consistent speed variability be maintained throughout the run?
First and foremost, pre-training your animals will eliminate many issues. This is not limited to gait analysis: When I was a graduate student, we always pre-trained our animals before all maze testing.
That said, CatWalk XT has two ways to maintain speed variability. First, the camera is pointed to an area on the walkway that does not include the beginning or end. This allows the animal enough time and space to initiate movement and maintain a stride. Second, CatWalk XT has a setting that requires the animal to maintain a consistent speed in order for that run to be compliant. This means that the animal cannot go slow and then speed up and then slow and then speed up. It has to be consistent throughout each run.
Could you please tell us about CatWalk XT?
CatWalk was originally designed by Frank Hamers in the early 2000s. Noldus officially released CatWalk in 2006. The original design used white LEDs for the walkway, had an analog camera (i.e., ~30 fps), and no illuminated ceiling.
In 2009, Noldus released CatWalk XT. The walkway had green LEDs, a new ceiling which was illuminated with red LEDs, the camera was updated to digital and high speed (100 fps), and the redesigned software automatically labeled, classified, and analyzed the footprints. Since then, there have been over 2,400 publications using CatWalk XT, six times more than any other gait system available.
The turnkey system comprises the frame on wheels, including the camera, walkway, ceiling, control unit, etc, and the computer with software. As I mentioned, CatWalk XT works by automatically detecting the contact of the paw on the glass surface. The paw reflects the light towards the camera. Therefore, CatWalk XT detects the footprints based on the animal and not by an algorithmic assumption, as is the case with some other systems.
Thanks to Illuminated Footprint technology, CatWalk XT takes the green signals from the camera and automatically measures the size of paw, its position on the walkway and its intensity. From there, the software automatically calculates the relational and temporal footprint information into over fifty parameters, along with both two- and three-dimensional visualizations.
How does the system assign values to the pressure of the animal walking throughout the walkway?
It is important to note that CatWalk XT measures intensity of light and not pressure. The light intensity is measured on grey scale value of arbitrary units, or 0 – 255 AUs. The Detection Settings within the software calibrate the green and red light voltages to ensure equal paw intensity throughout the experiment.
How is that data – the footprints – transformed into information?
The Illuminated Footprint technology works like this: Green LED light is internally reflected inside the glass walkway. Light bounces between the glass-air boundaries because the reflective index of glass is higher than that of air. The contact of a paw on the glass surface allows light to escape because the paws have a higher reflective index than the glass. The paw reflects the light towards the camera underneath the glass plate.
CatWalk XT automatically detects the green light ‘blobs’ (paws) in the video. The Automatic Footprint Classification Module assigns the blogs as paws (right front, left hind, etc.) based on the direction of movement and the type of rodent (rat vs. mouse). The surface area and XY position is calculated 100 times per second.
The animal’s body contour is created by the red-illuminated ceiling. From there, CatWalk XT automatically creates relational footprint information, such as stride length, the base of support, and print positions, along with temporal footprint information, such as cadence, step sequence, regularity index, phase dispersions, stand, swing, max contact, etc.
Can you explain a little bit more about data visualization?
As noted in the literature, one distinct advantage of CatWalk XT is the data visualizations. First and foremost, you have to run the video itself. This provides a replay of the animal’s run with the classified footprints at an adjustable speed (1/25, 1/10, 1/5, ½, or 1). In addition, three separate visualizations show the paw and run information in a two or three-dimensional plane.
The 3D Footprint Intensity shows each paw by the print length, print width, and intensity. These graphs are moveable, so you can see every aspect of the paw’s interaction with the glass plate. The 2D Footprint Intensity shows the same information but it is graphed over time. Each paw is labeled with a unique color showing the maximum and mean intensities over the course of any given run.
Finally, the Footfall Patterns display the type of pattern used by the animal (i.e., cruciate, alternate, or rotating). These visualizations may be exported as image files or along with the run video itself into a separate *.avi file. These are perfect for lab meetings, presentations, and defenses.
Can you tell us more about the parameters involved in CatWalk?
I have mentioned some of the parameters already, but one advantage of CatWalk XT is that each parameter is described in full in the manual. In many cases, the exact formula is shown along with a publication to support it. CatWalk XT provides data on the individual paws, the single runs, and overall trials.
The software allows you to choose the parameters most relevant for your research. All of these parameters are calculated per paw, as well as the front and back paws: initial stand time, stand time, stand index, max contact (time, area, intensity), X & Y position, print length, print width, print area, mean intensity, swing speed, stride length, step cycle, duty cycle, phase dispersions, the base of support, cadence, step sequence, percent time standing on a single paw, two paws (diagonal, girdle, or lateral), three paws, and four paws.
If you have the Interactive Footprint Classification Module, you get the Sciatic Functional Index, Peroneal Functional Index, Posterior Tibial Functional Index, toe spread, intermediate toe spread, paw angle, etc.
Could you tell us about RunwayScan and whether it is a good alternative to the CatWalk system?
As I mentioned with some other methodologies, I have never used or seen RunwayScan. The data I have is based on published literature or from the company itself. That said, RunwayScan also uses a high-speed camera, unforced free walking, and natural movement, and it also provides over fifty gait parameters.
However, there are a few differences. For example, there is not a lot of information publicly available, and there are few citations in the literature. Furthermore, there is no visualization within the software. However, you do get the same print length, toe spread, gait angle, stride frequency, stride length – so a fair amount of data is recorded.
What is the classic rodent gait measuring system ‘Ink and Paper’?
The standard and the most classic way to go about it is with ink and paper. This is super simple and very cheap. The idea is that you put ink on the animal’s paws and let it run across a piece of paper. This originated before we had any automated systems.
However, there are a number of drawbacks to this method. As you might expect, it is messy. However, the main difference is that everything is manual. The user has to classify the paws and all relational footprint information. In addition, no timing information is generated, such as swing speed, run speed, etc., and it is exceptionally difficult to measure intensity or pressure. Finally, the data is difficult to interpret. This system is one of the reasons why automated systems have come about.
What trends have you observed in citations for CatWalk, DigiGait and TreadScan in terms of publications and reports?
As I mentioned previously, CatWalk XT has six times the number of publications than any other system. Over the last five or six years, researchers publishing in the top neuroscience journals (Nature, Neuron, Science, etc.), are more likely to use CatWalk XT. According to Google Scholar, there are about twice as many publications for DigiGait as there are for TreadScan, and there are about twice as many CatWalk publications as there are for DigiGait. Again, CatWalk XT has been used in over 2,400 publications, which is why we feel it is the best in the market today.
What factors should influence the choice of methodology or system?
The choice of methodology in the system really depends on your experimental questions and needs. I have spoken with researchers who must use a treadmill system. Others find that ink and paper fit their budgets. What matters most is that researchers get exactly what they need in terms of parameters.
Furthermore, the system should reduce the stress on the animal and generate data that is reliable enough to publish and justify the use of the animals.
What is important when having conversations about what systems are right for the user?
We believe that conversations should be about what is best for you as the researcher. It should not be about which company is better or worse, but instead should be about what system is best for you, your science, your question, your space, and your availability.
We need open conversations about these types of methodologies and systems so that we can best serve the research. Noldus’ core values are building lasting relationships, operating with integrity and transparency, and contributing responsibly. We can all get better at providing what is ultimately important, which is providing quality images and data, reducing the harm to the animals, reducing the number of animals required, and providing the best quality data possible.
Can CatWalk data be used to boost sample size analysis with DigiGait?
My gut reaction was that it is apples and oranges because we are looking at the treadmill system versus a non-treadmill system, but that is my initial reaction. There are some publications that compare the two systems and show how to use the DigiGait data to look more like CatWalk. But even in the human world, there is plenty of data to show that running on a treadmill does not match the data generated when free running. That is to say that there are changes to the gait and the overall assumptions about how they work. In many cases, their applicability does not necessarily match. The answer is somewhere between a yes and no, maybe, depending on the research question and exactly what you are getting out of it.
About Jason Rogers, Ph.D.
Jason Rogers, Ph.D., is a Senior Research Scientist at Noldus Information Technology. He has a doctorate in psychology with a focus on the neurobiology of learning and memory. Jason joined Noldus in 2008 as Application Specialist and lead trainer. Since then, he has given hundreds of training courses and presentations around the world.
About Noldus Information Technology
Noldus Information Technology was established in 1989 by Lucas Noldus, founder and CEO of the company. With a Ph.D. in animal behavior from Wageningen University, he developed the company’s first software tool during his research in entomology. Noldus has strived to advance behavioral research ever since, evolving into a company that provides integrated systems including software, hardware, and services.
We now offer a wide range of solutions for research in animal and human domains, including biology, psychology, marketing, human factors, and healthcare. We work with leading suppliers and develop innovative, state-of-the art products. We also offer excellent technical support and customer care. As a result, our systems have found their way into more than 10,200 universities, research institutes, and companies in almost 100 countries.
The success of our company is determined to a large extent by the enthusiasm and creativity of our employees. We encourage each other to think outside the box, which leads to unique products and services for our customers. And we are always on the lookout for new talent!
Noldus Information Technology