I'm Doctor Matthew Sinnovich, and this is a webinar on the review of inertial census in the equine lameness exam. To start off, I'd just like to state that I have no conflicts of interest to declare. This talk is not sponsored in any way, or by any particular company.

All opinions expressed here are my own. Many of the examples will be based on the system that I currently use, but there is no clear benefit to one system over the other at present, and their use should be determined by the clinician's own practise should you decide to use them. And as I said, opinions are all of my own.

Right. So lameness is a challenge for many equine clinicians. Inertial sensors offer us the potential of an unbiased assessment of the horse's gait.

And hopefully in this webinar, what we'll do is look at a couple of the different systems, get into some of their uses and some of their limitations as well. Now, the learning objectives are listed there, and you can have a read through those. Initial sensor systems, for the objective analysis of gait asymmetry were initially only really available as a research tool in, orthopaedic and movement laboratories such as the RVC, and these we use often alongside force plates and retro reflective video capture systems, which are still some of the state of the art, with regards to research and Determination of asymmetry in horse movement, particularly with regards to research publications.

However, these have now, the inertial sensor systems certainly have become small and portable, and their commercial availability has meant that they have become very accessible to private practitioners. This has meant that their permeation into the market, is strongly evidenced and this can be seen by the large and very increasing number of research papers produced using this technology as an objective reference. Now, from a point of view of objectifying the lameness exam, that's a very good thing for research and research publications.

But this has also sparked quite a big debate in the scientific community as to what the exact definition of lameness is, how we should be recognising it, and also how we should be measuring it. Recently, there was an ongoing and informative discussion by expert diagnosticians on lameness, and this has been presented essentially through the editorial content of the equine vet Journal. One of the big controversies controversies focused on just exactly what is lameness.

One concern is that attaching a label of lame to a horses fought with many negative connotations. This may impact both horse and owners perceptions of performance, their fitness to compete, have impercussions or repercussions on their sale. And another concern centres on whether lameness is a disease in itself or simply a terminology that encompasses a number of clinical signs representative of some other underlying pathology.

Most practitioners will likely agree just that as colic is not a disease in itself, but it's rather a descriptive term for abdominal pain that develops to a number of other reasons. Similarly, lameness may be considered descriptor that indicates that the horse has asymmetry of bilateral movement. Now this may or may not be related to pain and may describe bilateral movement as the comparison right to left parts of stride as a way to shift it to a less painful limb.

Another concern expressed by these editorials is whether all gait asymmetries represent some abnormal problem in the horse. Some clinicians have pointed out that not all gait asymmetries correlate to an underlying pathological problem or necessarily pain. Some asymmetries are just a mechanical or neurological issue or may just represent the way the horse physically moves.

In addition, a horse that has bilateral pain, so is asymmetric in two sides, thus essentially rendering it symmetrical again, may be lame and have pain, where both limbs experience pathology. But they, that may be mixed, missed, in being picked up by an inertial sensor system. So what are inertial sensor systems?

They use body mounted inertial sensors, accelerometers, GPSs, and gyroscopes, depending on the system, which are situated at specific points on the horse. And these then collect data in real time about pitch, roll, and asymmetry as the horse moves. This is then transmitted through a receiver, to a processor, usually a computer or a tablet where software then uses algorithms to analyse and record this data.

This is then processed through various means into a report for the veterinarian to interpret themselves. Some systems also have an interpretation function. This will then use the data received and translate this to a quantifiable asymmetry and then hopefully through that determine the lame limb or limbs.

This allows objective gait analysis to be formed essentially in real time and in the real world. And this combined with a portability has meant that these systems have the potential to increase and aid the determination of subtle as well as complex lamenesses encountered by equine clinicians. In an ambulatory setting as well as hospital settings.

One of the biggest perceived advantages of the use of objective gate analysis in the layness evaluation is the possibility to remain impartial and to remove the bias from both the original interpretation of the lameness, and also to remove bias from the outcome of perineural or regional anaesthesia, because these changes can now be physically quantified, and a value put on them to see whether your nerve block is working or not working. So inertial sensor systems measure the two main indicators of lameness that are seen visually during a lameness exam. And what these look like is hip hike and head nod.

Vets with large experience in lameness evaluation are more likely to use many other movement parameters in conjunction with these in their examination. And this has been proved through a number of research papers and validation papers. And this is where some of the debates around the use and interpretation of inertial sensor systems has come in.

So typically in the four limb lameness, what we look at, or certainly where we start looking and where the system start to measure is in the typical head nod. And this is caused by the horse shifting its weight towards its hind quarter away from the lame limb. It tends to be louder on the opposite limb, so especially if you're listening on a or trotting on a tar or a hard surface, and this tends to more easily be seen when coming towards you.

Hind leg lameness is typically we talk about the that hip hike, which is caused by the horse shifting the pelvis to accommodate the leg movement and weight bearing during pain. The lame side tends to be the one that moves through a greater range of motion. And here, when we are viewing the lameness, we tend to look at the tumour coxa.

Now this is often best seen with the horse going away from you, usually in the trot. And can often be visualised as reluctance to flex the joints in the hind limb. And this is equivalent to a tilt on the affected side of the pelvis up as the leg is moved forward.

Looking from the side, you need to fully assess limb function, and this is where you would see an arc of foot flight which may be symmetrical or asymmetric. Now, typically, this would be what you would see in the hip hike. That's quite an exaggerated video over there, but just for demonstration purposes.

So what are the inertial sensor systems measuring? Essentially, what they do is algorithms calculate the head and pelvic height differences. And this is measured in millimetres between the right and left halves of the stride.

And then sophisticated proprietary algorithms convert the acceleration, which is measured by the head and the pelvic accelerometers to the relative position of the head and the pelvis throughout the stride cycle. So in the trotting horse, every stride cycle consists of 2 maximum head positions and 2 minimum positions in the head, and a single up and down in the pelvis, one for each side. Thresholds for mean height differences are then measured between the right and the left halves of the stride.

Now for some systems, these are sort of within about 6 millimetres for the head and about 3 millimetres for pelvic movement. These height differences are referred to as the differential max and differential minimum of both the head and the pelvis. A large body of scientific data exists, which has been used to validate these systems, and they have been put through tests under all sorts of circumstances.

They're validated on different surfaces for different types of movement, . And have also been shown that they. The collection can be affected by incorrect placement of sensors, changing variations and things in stride, horses playing up and being too excitable.

And this is one of the very important parts about using these sorts of systems is that you need what we call a stable lameness. So a horse needs to have been trusted a couple of times for the layers to have settled, for their excitation to have been blown off, if they're throwing their heads around or being stupid or anything like that, often they need to settle before the proper examination can take place. And this has been shown through the validation, references.

So here we have two systems, the aggregate system on the left there, and the equinosis with lameness locator or equinosis Q on the right. As you can see, a number of different systems have a number of different sensors. The Equigate system comes with different options of A number of different sensors.

So you can have a group of 6 or a group of 4, which can then be placed in different positions depending on how you are assessing your horse. The equinosis on the right there is a very, very simplified version and just has 3 sensors. They're well labelled, and fairly foolproof to use.

The head sensor there is a uni axial accelerometer. The right forum sensor is a uni axial gyroscope. And then the pelvic sensor is a uniaxial accelerometer, although lani's locator only uses the vertical acceleration collection of their data.

Over here is a slide just demonstrating how the horse gets set up with the instru instrumentation. And you can see it comes with nice nearrene head pole guard and a passion guard, and these sensors just stick in there with some Velcro. There is a separate little cage that then goes on the pelvis, and it is all fed through to a customised tablet.

As we said before, it's very important to have suitable suitable facilities. Now, this is true of any lameness exam and not necessarily just using inertial sensor systems. Very important there to have a nice even placed trot.

As you can see, you wouldn't want to be doing a lameness exam in the bottom right field. You wouldn't be able to see anything, and I think even an inertial sensor system is not going to be able to help you out of your problems if you're, if you're sitting in that. So it's advisable always to perform two straight line trials, one after the other to confirm the results.

And if a lameness is present, you need a stable lameness. So that is going to be able to be seen, consistently and not be dependent on one specific surface or direction. Surface is always an option and this can be added in the straight line trials and so, sometimes these sensors will take into account different surface types.

For lunging trials, a surface should always be chosen. You want to collect at least 25 total stride lengths for a straight line trial, and at least 6 continuous strides before the horse stops and turns around. If performing a lunging trial, you definitely want to collect about 2 screens' worth of data, and that will be equivalent to about 45 to 50 strides worth of data, which will give you enough of a stable data and enough of a data set for the algorithms to then calculate what they need to do and to extract the best value of the data out of that system.

This is what the raw data essentially looks like. This is a nice stable collection. You can see it's a nice even bar the whole way across through both.

The white bits in the centre represent the point where the horse slowed down, turned around, and then trotted back. And for the lameless locator, this or the equinosis, this is basically what your report would churn out. So you have your trial metadata, your owner, your horse, specific trial data, including time, date, type of trial.

And then you would be able to have a look at your differential loading assessment. So on the left there, you'll have a ray diagram plot, and these are graphical representations of the trial data. And then you have some calculated variables underneath there.

These are then more specific about the quantitative data from the collective trial. The calculated lameness variables include the mean values on the diff max, and the diff min of the head. So that's the highest point and lowest point of the head, and then the diff max pelvis and diff min pelvis.

So the highest and lowest point of the pelvis. And then standard. Deviations for each of the respective means in the diff max and the dmin.

The importance of that is that you can see that your standard deviations are within a recognised limit. So according to basic statistics, you know that the quality of your data is good enough for the trial, and it means that you then have A good reliability in the data that you have. And then you also get the total difference between the head and the mean vector sum of the forum data.

You'll be able to see how many strides a selection were selected. So the blue column over there will show you which set of data was taken in the interpretation. And then in some systems, you get an automated interpretation, with a degree of evidence.

So that would be that corner in the bottom there, which is called the aid. And then there's some evaluator notes so you can type in your own notes there as you're going through the, through the exam. So, basically, what we're looking at as we go through the data is the diff max of the head.

Now, this is the difference in the maximum head height that occurs before the right front weight bearing compared to before the left front weight bear. This is measured in millimetres and is reported as a mean value of all strides. The diff max head equals the diff the maximum height of the head before right front weight bearing minus the maximum height of the head before left weight-bearing.

The difference of the head is the difference in minimum head height that occurs during right fronts mid stance, compared to during left front mid stance. And that is also measured in millimetres and reported as a mean value of all of the strides. And then again, the diffman head equals the minimum height of the head during the right front midstance minus the minimum height of the head during the left front midstance.

The diff max and the diff min of the head are intimate related and while measured separately, they must be considered together in the determination of the side and timing. So whether it's an impact, a mid stance or a push off type lameness, The diff max and diff men of the head may also be positive or negative values depending on the side and the timing of the lameness. Now what this looks like over here.

Yeah, so here are a couple of examples of movement patterns in a right for lameness. So the blue bar there is the indication of the right r stance phase, and you get the values there with your normal variation between -6 and 6 millimetres. And as you can see there in the second one across, you're having a change in those values.

So you would have a right front impact lameness over there. Third one down is an indication where you would have a right front mid stance lameness and the lower one would be an indication of what a push off type lameness might look like. Heading towards the back end, the diff max of the pelvis is the difference in maximum pelvis height that occurs before right hind weight bearing compared to before left hind weight bearing.

Again measured in millimetres, reported as a mean valuable strides. Right hind asymmetries are assigned a positive value, left hind asymmetries are assigned a negative value. Elevated diff max are either positive or negative, then correlates to a push of lameness in the hind limb.

With the diff max of the pelvis equaling the height of the pelvis before the right hind weight bearing minus the maximum height of the pelvis before the left hind weight bearing. The minimum over there again, then is the difference in minimum pelvic height that occurs during right hind mid stance compared to during left hind mid stance, measured again in millimetres and reported as the mean value of all strides. The right hind asymmetries are assigned positive values and the left hind asymmetry is negative values.

And then an elevated dmin, it's either positive or negative depending on which side, correlates to an impact lameness in the hind leg. So contrary to the determination of the four laminates, which depends on the relationship of both the diff max and the din of the head, the diff max and diffin of the pelvis are independent variables. So if the horse has a push of lameness, the pelvis will rise less than it should when the affected limb pushes off.

Which will result in a difference in maximum heights between the right and left halves of the stride. If the horse has an impact lameness, the pelvis will fall less than it should on the affected sides and stance, and this will result in a difference in the minimum height of the pelvis between the right and left halves of the stride. In some cases, both an impact and push of lameness can be evident, and this will be shown.

So there we go, this just shows different variations over there with a right hind push off and impact versus a right hind impact or a left hand push off and right hind impact, and going through the different permutations of them over there. And this would show the example of the different pelvic movement patterns in a right hind. Lasus again, right hind stance phase is indicated by the light blue bar in the top graph over there.

And you can see the top is normal with the second one indicative of a right hind impact type lameness. The third one, a right hind push off type lameness. And the last one there, the right hind impact and push off type laus with the two different variables and the change in the curve.

Now, we'll just go through this fairly quickly, and this basically is just how to interpret the, the data as you're kind of going along. Where you want to start with your interpretation is to inspect the ray diagram plots. So you're gonna have a look there to see if there's an accumulation of rays in any particular quadrant.

You're gonna have a look at the summary of the ray plots, both for the front and the hind end. Then for the forumb, you're gonna inspect the total diff. Head, so what you're looking at there is the mean vector sun.

And that was, represented there in the red bar. And that, as you can see, there's a graph underneath it of an impact la it's just showing the vector sum, which will be an average of those vectors through there. And then inspect the diff max and diff min values for the pelvis, which will be shown by that bar graph graph on the right.

And as you can see that Then is broken up into push off and impact sides, both left and right. So you go through all of that and see where the errors occurred, and that would then be able to lead you to an interpretation of which side was lame, how lame it was, and, what sort of compensatory patterns are going on. Now, only really a veterinarian can confirm whether the horse has a compensatory lameist or two separate laisfoy.

There are suggestions offered here as a general guide. They can be useful when the compensatory movement may be greater than the primary lames, and certainly that can happen and has been shown by a number of papers again produced through the use of different, systems. Again, this highlights the importance of going through a full lameness exam with palpation, lunging, flexing tests and blocking, which can help sort a primary form of compensatory lameness and also just help you show what the lameness looks like on different surfaces under different conditions.

Now, primary lamenesses in the forelimb can cause compensatory or essentially false vertical pelvic movement asymmetry, and primary lameness in the hind limb can also cause compensatory or false vertical head movement asymmetry. So knowledge of some of these patterns will assist the user in correctly identifying the locations of a primary lameness. The equinosis group have come up with what they call the law of sights.

And according to the law of science, a horse with a primary hind limb lamus may show compensatory ipsilateral forumb asymmetry, and a primary forelimblaus may show compensatory compensatory contralateral hind limb asymmetry. And this has been tested and reported in a number of peer-reviewed studies. So a primary for lamb lameness of any timing with impact push off, etc.

May cause a comp compensatory hind limb lameness push off asymmetry. So for example, here, the law of sides would say that a right front lameness with a compensatory left hind push off lameness. And there are various permutations of this, as you can see, going through the, the descriptors over there.

The important thing I think here is to understand the system that you're using, understand its limitations. This is pure data. It just records asymmetry and chucks the values back out to you.

It's very important for the clinician to know what they're talking about and to know their system and the limitations of their system, and then be able to interpret the data accordingly. So here we have an example with a left hind lameness, of primary concern with a compensatory left front asymmetry on the left, and then on the right there, you'll see the data collection for a primary right front lameness with compensatory left hind push off asymmetry. You can see the red bars, extending well into the abnormal range there.

And on the right front impact, you can see that nice big red vector sum, through the centre. Here we have just a quick reference. So there's various permutations of for limb lameness and where the vector sums may end up.

And the same thing is available for the hind end. And basically, if you have one of these systems, the best way to do this is to do numbers and numbers of horses to put it on every single lameness that you see and every single lameness that you do, . Look at the lameness independently, so see where your head is at and then see if the data correlates to it.

Now the data is not always going to correlate exactly to what you're seeing, but it can always be worked out and teased out. And I think that's one of the very important things with these systems is that they're not there as a complete answer. They are like an X-ray, like an ultrasound.

They only report what they see, and it's up to the clinician to determine what that means and how it fits into the horse and the patient that is sat before them. So do we need them, if it's still up to the clinician to determine the ultimate outcome and to syphon through all of the data and pick out the bits that is useful, and to then use that to determine where they go to. As with all scientific tests, due to differences in sensitivity and specificity, there will always be the possible possibility of false positives or negative.

And they do exist a population of horses that are seen by their owners in sound, and who will fall within the thresholds of lameness, according to an objective gait analysis. And this has sparked a debate about whether there's a definition of lameness that needs to be revisited in the age of objective gait analysis. It's been hypothesised that there will be a larger proportion of laying horses in the group, that exceed the threshold than in a group that do not.

Bilaterally laying horses may also not exceed the thresholds of Objective gate analysis systems. And it's been found that the confidence intervals recorded on objective gait analysis is follow the threshold values and horses that were repeatedly examined with inertial sensor systems. And this is useful when determining the response to perineural analgesia when a positive block is likely if a change in value is obtained that is greater than the confidence in.

There's also a difference between lameness and asymmetry, and the two are not necessarily always the same. Lameness cannot also be described as simply lame or not lame, as various degrees occur across a continuum and various situations are possible. As such, a simple defined point of lame or not lame does not cover all eventualities.

And this is where the critical analysis provided by an experienced clinician is really required. The benefits of an objective gait analysis may however give the user some confidence in the results of their nerve blocks, particularly during poor performance workups or with low level asymmetries. That may be only be causing a reduced performance when no overt lameness may be visible, but improvement may may be reported by an experienced rider.

If an experienced rider or handle is not available. Objective gait analysis are often more sensitive than the human eye with respect to temporal and spatial resolution. And so these can often detect significantly smaller variations in symmetry than, clinicians may see visually.

It may also give clinicians more confidence in interpretation of the primary cause of a multi-limbed lameness and the interpretation of perineural analgesia and the examination of these types of autism and poor performances. So even though lameness is a continuum, objective measurement of the degree of lameness will improve the accuracy of interventions and the diagnosis ultimately related to perineural analgesia, as well as therapeutic interventions. Although there is debate about the definition of lameness and the value applied to the threshold values used in these objective gate analysis and systems, and the asymmetry and the interpretation of the lameness.

Everyone does seem to agree that objective gait analysis is a clinical tool that does not preclude a thorough clinical examination, and is only a means to aid in clinical decision making. So is their use inescapable? I think the use of activity and health trackers, pedometers, smartwatches has become increasingly popular amongst the general population.

Many devices are readily convercially available now. This uptake of technology started to cause quite an acceptance and a change in the mindset of many people about gate research, and this is, I mean, not just in the equine industry, but hugely through people and, and the medicine industry as well. New developments have seen motion tracking technology moving into Ferrari, where there's gate mapping software, and this is used to distinguish landing stance and breakover in in real time, as well as movement.

This is the Verkman 2020 and you'll see that in the next slide. This is alongside things like heart rate monitors, ECG monitors, which are now used to track equine cardiac performance, so like the polar systems, which are very useful in the hunting community and very useful for the racing community as well. Currently, there are only really 2 commercially available equine gate analysis systems.

And as we stated, these are the Eregate system and the equinosis Q. Unfortunately, for the moment, they are only available to veterinary clinicians, although I do foresee a time where systems like these can and or may become commercially available to be used by, Owners themselves, in the determination of monitoring their horses, and keeping an idea of where they're at from a day to day point of view. This here is just taken from the Verkman Black system, which is an example of inertial sensor technology that's used in Ferry.

And here it's used to map footfall and foot flight. Can also be used to check the stance phase and the breakover, and these can be measured using the system and essentially real time changes can be made to hoove shape, in order to improve breakover, swing and landing. A recent prospective qualitative research study was performed using a questionnaire to understand how existing movement analysis practises were being carried out, as well as perspectives from a variety of stakeholders in the Irish equine industry.

This is through the Thoroughbred racing, and elite sports horse producers. And this was focused on the use of objective gait analysis in the field. The authors reported 4 main themes after analysis of the data.

These were the importance of a tacit knowledge and experience, and in the holistic management of the equine athlete, management of the equine athlete is complex and requires a multi-layered approach to problem solving, as the perfect athlete does not exist. There was scepticism of the value of the technologies available and an awareness of their existence. But at the same time, the use and finally the key barrier to technology adoption was the economic value of the horse and the cost of implementation on a large scale.

Now, the objective nature of inertial sensor systems as a diagnostic aid, as well as the fact that clinicians can record and retain large amounts of data, negate many of the biases associated with mental retention and idollases. As the use of these sensor systems increases, so does the amount of data collected and the ability to interpret this data and to apply it to specific phases of stride affected in time, and this may be able to correlate to specific pathologies. Ultimately, this may then be able to enable targeted diagnostic anaesthesia and ultimately reduce the number of needle, sticks and time required in lameness exams.

This in turn may improve clinicians' safety by limiting the number of dangerous procedures required, especially when dealing with hind leg lamenesses and poor performances. The argument has also been made that inertial sensor systems cannot lie and that they only record the data they are programmed to, and as such their value is only as good as the interpretation made by the commission, and I, I stand by that. I think that's a, a very valid comment.

And the adoption of this and other technologies has been summarised by, Professor Derek Nottenbelt in the quote, Technology won't replace vets, but vets who use technology logically and carefully will replace those who don't. And I think that's, that's quite a good quote there by Professor Derek. So according to the most widely applied model of technological use and acceptance, the technology acceptance model, the two main criteria that predict behaviour and the use of And the use of the perceived of use and the perceived usefulness of the technology.

And trusting the data gathered and having faith in the interpretation of the data has been described as one of the barriers to a large scale acceptance of use of objective gait analysis in sectors of the equestrian owning population. There's a growing body of evidence for the use of these inertial sensors and their value in both research as well as in the clinical setting. Undoubtedly more and more research will be produced that will strengthen the case for the use of these inertial sensor systems.

So should you get one? As we said, it does not replace the clinical exam. It is an aid, not an answer, and one does need to understand the system and its limitations.

Concerns do exist that the technology is used at the expense of a thorough clinical exam, and that clinicians may develop a dependence on the technology and not further their own diagnostic skills. But suc successful integration of the technology has been shown to enhance productivity. But failure of a system can also cause over overall dis dissatisfaction as well as economic loss.

But with the pace of technological developments and an adoption of the 21st century, it would be difficult to argue that the use of inertial sensor systems and objective gate analysis would not become routine in the near future. Whether using existing systems or improvements thereof that address the owners and other stakeholders' concerns. So I think these are coming whether we like it or not, but, The important thing I think is to still develop your skills, to still know what you're doing with regards to a lameness exam and to understand the system that you're using because it will only ultimately be as good as the clinic operating it.

There are some references, and thank you very much for your attention.