001: Long Pastern Bone
The Long Pastern bone of the horse (aka "P1") is not symmetric -- it has two non-zero angular parameters that we will call "Tilt" and "Twist". The "Tilt" is the angular misalignment between the pastern-joint-axis and the fetlock-joint-axis as seen in a DP radiograph. The "Twist" is an angular twist about the long axis of the bone. One way to notice the Twist: place the bone on a flat surface - it won't lie flat - you can rock it, due to the Twist angle. In our measures, the typical P1 bone has a Tilt of about 3 degrees and a Twist of about 6 degrees. See 'The Hoof' book for more.
002: Computing the Axis of Rotation from Bone Shape
We are developing what we call "The Bone Engine" which is a collection of software tools that can do various computations on scanned 3-D polygonal models of bones and other structures. As an example of one ability of the Bone Engine -- it can compute the best approximation of the 3-D location of the axis of rotation for a given joint. The interesting thing about this method is that the computation is made purely from the 3-D shape of the condyles and epicondyles of the two bones that form the joint. The resulting location of the axis of rotation is derived purely from shape (not from motion data, nor someone's opinon, nor other means). The 3 three video segments show how the axis of the fetlock joint of a horse's front leg was determined from laser-scanned 3-D models of the long pastern and cannon bones.
003: Motion Analysis via Drone
An HD video camera on a drone, and a horse that is not spooked by it...
004: Robustness of the Palmar Angle Measure
Metron software measures the 'Palmar Angle' on a lateral hoof radiograph (among other measures). For a given x-ray generator, and using the Metron Block on a flat floor, the main source of misalignment is due to alignment of the block + hoof when the image is taken. But in our study (shown here) we find the measurement to be fairly robust: as long as the hoof is properly aligned within plus-or-minus 10 degrees, the Palmar Angle measure is accurate to within 2/3 of a degree. In the video below, we go all the way to plus-or-minus 15 degrees of misalignment. By the way, it is pretty easy to achieve plus-or-minus 5 degrees just by eye-ball -- hence, we think it is entirely practical to get a Palmar Angle measurement probably within a half of a degree. Another interesting thing about this video: the points placed on the image to measure the Palmar Angle were placed automatically by Metron's Intellect Module (see Deep Learning Technology) and I think you may agree that over this range of alignment, the points were positioned as well (or better) than an expert human user would. You probably will want to replay this video a few times to examine everything that is going on.
In the next video (below) we show the effect of changing the height of the block that the horse stands on. The red cross shows the location of the central x-ray beam as the height of the block is varied. At one extreme, the beam points across the top surface of the block, at the other extreme (3" higher) the beam points close to the location of the coffin-joint's center of rotation. Watch the metal markers built into the block and you can see the effect. The measured Palmar Angle is 5.1 degrees plus-or-minus 0.55 degrees. Again, we think this shows that the measurement of the Palmar Angle is fairly robust to this change in imaging geometry.
Another thing that might affect measurement of the Palmar Angle: the horse shifts its weight around and may load (and thereby compress) one hoof more than the opposite -- and might this change the Palmar Angle? The images below are of the same hoof and taken within a few minutes of each other. In one extreme case, all the weight was on the leg being imaged. Again we see that Palmar Angle measurement is fairly robust, as it changes only by a half a degree or so -- meanwhile the coffin-joint angle has changed by over 26 degrees!
Check back for additional postings in the Nerd Corner!