|
Post by dubiousgolfer on Mar 28, 2024 11:51:28 GMT -5
Dr Mann
Have you accounted for the fact that AMG are using thoracic 'rotation/side-bend/ext-flex' measurements relative to the pelvis? Their up/down measurement of the chest/pelvis seems contentious when looking at the GIF you posted.
With regards the thorax rotation measurements , GEARS is using the same AMM method where the thorax is the centre of a line drawn between the AC joints. For actual rotation of the thorax , they would have to measure it relative to the pelvis rotation.
DG
|
|
|
Post by imperfectgolfer on Mar 28, 2024 12:59:42 GMT -5
Dr Mann Have you accounted for the fact that AMG are using thoracic 'rotation/side-bend/ext-flex' measurements relative to the pelvis? Their up/down measurement of the chest/pelvis seems contentious when looking at the GIF you posted. With regards the thorax rotation measurements , GEARS is using the same AMM method where the thorax is the centre of a line drawn between the AC joints. For actual rotation of the thorax , they would have to measure it relative to the pelvis rotation. DG I am presuming that the pelvis is neutral and that their spinal rotation measurements are only related to the T-spine. If GEARS is using a line drawn between the AC joints it would be affected by scapular motion (retraction/protraction) and I presume that they must be using a line drawn across the rib cage at that level. However, as I stated when looking at RM at P5 in the capture image below, if one drew a line across the front of his chest at the level of the AC joints, it would not be ~60 degrees rotated clockwise relative to the pelvis (which is square at P5). I have made a new overhead image of RM at P5 (note that his lead arm is angled back slightly so that his hands are opposite his right upper chest as seen in MG's P5-image).
I have drawn black lines across his anterior/posterior chest at the level of his AC joints. That line measures ~35 degrees relative to his squared pelvis. Jeff.
|
|
|
Post by dubiousgolfer on Mar 28, 2024 19:04:13 GMT -5
Dr Mann I'm assuming the 58.7 degree shoulder rotation relative to the pelvis (at P5) in the AMG video is called 'X-factor', as described in this you-tube video by Shaun Webb. Here are Rory's AMM3D graphs: The white line is the difference between the thorax (pink line) and Pelvis (yellow line) measurements (ie. X-factor). I wouldn't know why your measurements differ from the AMM3D but the view angle you are using to compare thorax vs Pelvis rotation angles looks different to that used by Shaun Webb. GEARS view angle is normal to the horizontal plane and seems to be using projected angles onto the horizontal plane to measure 'X-FACTOR' , while your Rory image view angle is not normal to the horizontal plane. DG PS. What I find interesting, is that the X-factor increases slightly (maybe 10-15 degrees) from before P4 to maybe P4.5 . Does this imply that Rory's pelvis rotation is driving the upper body rotation?
|
|
|
Post by imperfectgolfer on Mar 28, 2024 21:48:13 GMT -5
Dr Mann I'm assuming the 58.7 degree shoulder rotation relative to the pelvis (at P5) in the AMG video is called 'X-factor', as described in this you-tube video by Shaun Webb. Here are Rory's AMM3D graphs: The white line is the difference between the thorax (pink line) and Pelvis (yellow line) measurements (ie. X-factor). I wouldn't know why your measurements differ from the AMM3D but the view angle you are using to compare thorax vs Pelvis rotation angles looks different to that used by Shaun Webb. GEARS view angle is normal to the horizontal plane and seems to be using projected angles onto the horizontal plane to measure 'X-FACTOR' , while your Rory image view angle is not normal to the horizontal plane. DG PS. What I find interesting, is that the X-factor increases slightly (maybe 10-15 degrees) from before P4 to maybe P4.5 . Does this imply that Rory's pelvis rotation is driving the upper body rotation? You wrote-: " I'm assuming the 58.7 degree shoulder rotation relative to the pelvis (at P5) in the AMG video is called 'X-factor'."No. In the latest AMG-video the 58.7 degrees of rotation at P5 refers to thoracic spinal rotation, and not to upper torso rotation (although they may be equivalent if the measurement of spinal rotation is simply derived from the upper torso's rotational measurement). If the pelvis rotates before the upper torso between P4 => P4.5 it creates dynamic X-factor by stretching the abdominal oblique muscles, which supposedly allows the upper torso to subsequently rotate faster.
The AMM graph suggests that the spinal rotation is >60 degrees between P4 => P5. I don't clearly understand how that it is biomechanically possible if the thoracic spine is flexing ~45 degrees between P4 => P5. It is hard for me to imagine that RM's spine is still rotated >60 degrees clockwise at P5 when looking at either overhead or back view images of his upper torso.
Jeff.
|
|
|
Post by dubiousgolfer on Mar 29, 2024 6:48:20 GMT -5
Dr Mann
I've been thinking about the AMG video where they are measuring spine motions relative to the pelvis. That means they are making a mathematical decision that the pelvis is an inert segment , while the spine is doing all the active measured movements. Obviously , that is not true in reality because the pelvis active movements could be changing the spine 'extension/flexion/side-bend/rotation' measurements.
For example , if at P4 , say Rory has 'x' degrees spine extension relative to the pelvis, then in reality the pelvis could laterally move away from the ball-target line while going into posterior tilt by P5 , which would decrease the measured spine extension from x -> y degrees (while the degree of rotation of the spine remains relatively unchanged relative to the pelvis). Therefore, it is the active pelvis movement creating the change in spine extension measurement , not any spine movement.
Further , say AMG mathematically made the centre of thorax the 'inert' segment , and then used the positions of the pelvis relative to the thorax to make measurements of spine extension/flexion/side-bend / rotation. Then wouldn't that be implying that the pelvis is the active segment causing all the change in measurement angles?
What I am saying is that their spine measurements of extension/flexion/ side-bend/ rotation relative to the pelvis is implying active muscular spine segment movement that may not be happening in reality.
Does that make sense?
Below is an issue that is very difficult to understand but is important to take into account when questioning the anatomical reality of 3D measurements.
A possible problem with 'AMM/GEARS' and other 3D systems' (too complicated to explain in any detail) is a mathematical issue regarding the order in which 3D systems are programmed to measure body segment positions/movements about the x,y,z axis. For example, AMM might measure a change in thorax 'position/orientation' from position A ->B using an ordered measurement 'ext/flex, rotation, side-bend' , whereas another 3D system might measure the change in thorax position using 'side-bend, rotation, ext/flex'.
This means that although the Thorax position/orientation has changed from position A -> B as measured by both systems , their measured values of extension/flexion, rotation, side-bend could be different.
This could theoretically mean different 3D systems providing different measurements of 'ext/flex , rotation, side-bend' for a particular segment change in position/orientation.
There is a research paper that explains this but unless you have some advanced level mathematical knowledge , it will be very difficult to comprehend. The thesis research paper is titled:
"An Investigation of Cardan Rotation Sequences on Lumbar Spine Lifting Kinematics"
Here are a few extracts from that thesis but you will need to understand Euler angles and Cardan sequences.
"Euler angles are commonly used for kinematic descriptions due to their ability to translate to clinical contexts and the conciseness of using three rotation angles. A challenge with using Euler angles is their dependency on Euler sequence selection. Many studies do not use the same Euler sequence, leading to difficulty comparing across studies."
"Differences in Euler/Cardan sequence have been shown to result in significantly different values of angular motion. Attempts to standardize Euler/Cardan sequences for different joints have been made, but the adoption of these standards has varied. Different definitions of coordinate axes also exist, creating further confusion and inconsistency among recommendations and studies."
DG
PS. Dr Kwon is actually collecting data for various 3D systems to ascertain what sequences they are using to measure segment kinematics. So I expect he will be producing a research paper about this in the near future.
|
|
|
Post by dubiousgolfer on Mar 29, 2024 8:58:24 GMT -5
Dr Mann
I've had to make frequent changes to my previous post because of the strange way AMM/GEARS define the spine segment.
DG
|
|
|
Post by imperfectgolfer on Mar 29, 2024 9:59:46 GMT -5
Dr Mann I've been thinking about the AMG video where they are measuring spine motions relative to the pelvis. That means they are making a mathematical decision that the pelvis is an inert segment , while the spine is doing all the active measured movements. Obviously , that is not true in reality because the pelvis active movements could be changing the spine 'extension/flexion/side-bend/rotation' measurements. For example , if at P4 , say Rory has 'x' degrees spine extension relative to the pelvis, then in reality the pelvis could laterally move away from the ball-target line while going into posterior tilt by P5 , which would decrease the measured spine extension from x -> y degrees (while the degree of rotation of the spine remains relatively unchanged relative to the pelvis). Therefore, it is the active pelvis movement creating the change in spine extension measurement , not any spine movement. Further , say AMG mathematically made the centre of thorax the 'inert' segment , and then used the positions of the pelvis relative to the thorax to make measurements of spine extension/flexion/side-bend / rotation. Then wouldn't that be implying that the pelvis is the active segment causing all the change in measurement angles? What I am saying is that their spine measurements of extension/flexion/ side-bend/ rotation relative to the pelvis is implying active muscular spine segment movement that may not be happening in reality. Does that make sense? Below is an issue that is very difficult to understand but is important to take into account when questioning the anatomical reality of 3D measurements. A possible problem with 'AMM/GEARS' and other 3D systems' (too complicated to explain in any detail) is a mathematical issue regarding the order in which 3D systems are programmed to measure body segment positions/movements about the x,y,z axis. For example, AMM might measure a change in thorax 'position/orientation' from position A ->B using an ordered measurement 'ext/flex, rotation, side-bend' , whereas another 3D system might measure the change in thorax position using 'side-bend, rotation, ext/flex'. This means that although the Thorax position/orientation has changed from position A -> B as measured by both systems , their measured values of extension/flexion, rotation, side-bend could be different. This could theoretically mean different 3D systems providing different measurements of 'ext/flex , rotation, side-bend' for a particular segment change in position/orientation.
There is a research paper that explains this but unless you have some advanced level mathematical knowledge , it will be very difficult to comprehend. The thesis research paper is titled: "An Investigation of Cardan Rotation Sequences on Lumbar Spine Lifting Kinematics"
Here are a few extracts from that thesis but you will need to understand Euler angles and Cardan sequences. "Euler angles are commonly used for kinematic descriptions due to their ability to translate to clinical contexts and the conciseness of using three rotation angles. A challenge with using Euler angles is their dependency on Euler sequence selection. Many studies do not use the same Euler sequence, leading to difficulty comparing across studies."
"Differences in Euler/Cardan sequence have been shown to result in significantly different values of angular motion. Attempts to standardize Euler/Cardan sequences for different joints have been made, but the adoption of these standards has varied. Different definitions of coordinate axes also exist, creating further confusion and inconsistency among recommendations and studies."
DG PS. Dr Kwon is actually collecting data for various 3D systems to ascertain what sequences they are using to measure segment kinematics. So I expect he will be producing a research paper about this in the near future. I cannot understand why the causal factors causing spinal movement will affect the accuracy of measurement of spinal rotation, presuming that the markers used for making the measurement are scientifically valid. Jeff.
|
|
|
Post by dubiousgolfer on Mar 30, 2024 5:51:27 GMT -5
Looking at AMG video again : Also, note what was said in the video (see frame images below), that the "spine isolated from the tilting and turning of his pelvis". But isn't the 'spine' defined as a line drawn from the centre of the thorax to the centre of the pelvis? Therefore wouldn't a lateral move of the pelvis away from the ball-target line (ie. no change in tilt/turn) , create a change in the extension/flexion angle? So , the pelvis might not be tilting and turning but could still be moving back laterally affecting the spine angles. The camera view could still be viewing perpendicular to the pelvis while all this is happening. So one cannot assume that Rory is physically muscularly flexing his spine +40 degrees from P4-P5 without taking into the account how the lateral movement of the pelvis could also be affecting that measurement. I've drawn the images of the defined spine line between centre of thorax to centre of spine . The pelvis moves laterally to the left while the camera is still viewing from a perpendicular standpoint. You can clearly see that the motion of the pelvis could also be causing a change in the spine flexion angle. A viewer of AMG's video might be confused and think he/she needs to try and make their spine flex like Rory from P4-P5 , but miss out on his pelvic movements that could also be contributing to that spine flexion. Further , here are some graphs I found on Phil Cheetham's blog showing the pelvis moving away from the ball-target line in the early-mid downswing (see top green line graph). I'm unsure whether this is a typical graph for tour pros in general. Is this also happening in Rory's downswing ? DG
|
|
|
Post by imperfectgolfer on Mar 30, 2024 9:24:34 GMT -5
Looking at AMG video again : Also, note what was said in the video (see frame images below), that the "spine isolated from the tilting and turning of his pelvis". But isn't the 'spine' defined as a line drawn from the centre of the thorax to the centre of the pelvis? Therefore wouldn't a lateral move of the pelvis away from the ball-target line (ie. no change in tilt/turn) , create a change in the extension/flexion angle? So , the pelvis might not be tilting and turning but could still be moving back laterally affecting the spine angles. The camera view could still be viewing perpendicular to the pelvis while all this is happening. So one cannot assume that Rory is physically muscularly flexing his spine +40 degrees from P4-P5 without taking into the account how the lateral movement of the pelvis could also be affecting that measurement. I've drawn the images of the defined spine line between centre of thorax to centre of spine . The pelvis moves laterally to the left while the camera is still viewing from a perpendicular standpoint. You can clearly see that the motion of the pelvis could also be causing a change in the spine flexion angle. A viewer of AMG's video might be confused and think he/she needs to try and make their spine flex like Rory from P4-P5 , but miss out on his pelvic movements that could also be contributing to that spine flexion. Further , here are some graphs I found on Phil Cheetham's blog showing the pelvis moving away from the ball-target line in the early-mid downswing (see top green line graph). I'm unsure whether this is a typical graph for tour pros in general. Is this also happening in Rory's downswing ? DG I do not believe that pelvic motion will affect measurements of thoracic spine flexion, rotation and side-bend. I also believe that those measurements only apply to the thoracic spine, and one can basically ignore the lumbar spine. Lateral pelvic motion may tilt the axis of the T-spine, but it should not affect the markers used to measure T-spine flexion/rotation/side-bend in such a manner that it would result in significantly inaccurate measurements. That dotted line reflects a changed degree of tilt of the T-spine, but not T-spine flexion which is happening roughly perpendicular to that dotted line. Jeff.
|
|
|
Post by dubiousgolfer on Mar 30, 2024 11:20:54 GMT -5
Dr Mann
I don't think those measurements are about the thoracic spine, they are about the Thoracic+ Lumbar spine (ie. the T spine).
Look at screen frame 10:26 in the AMG video.
DG
|
|
|
Post by imperfectgolfer on Mar 30, 2024 15:20:46 GMT -5
Dr Mann I don't think those measurements are about the thoracic spine, they are about the Thoracic+ Lumbar spine (ie. the T spine). Look at screen frame 10:26 in the AMG video. DG I agree that their spinal rotation/side-bend measurements include the lumbar spine, but virtually no spinal rotation or side-bend happens at lumbar spine level so it can be ignored. Jeff.
|
|
|
Post by imperfectgolfer on Mar 31, 2024 10:32:23 GMT -5
DG posted this image showing AMM-3D's measurements regarding thoracic spine rotation and side-bend and it shows that thoracic spinal rotation does not increase between P4 and P5 in Rory McIlroy's golf swing action and that side-bend temporarily increases to a small degree. These results are similar to MG's posted results (based on GEARS) so they are probably accurate. I have therefore spent more time thinking about what causes side-bend to temporarily increase at the start of the downswing and I have come up with a plausible explanation. Consider these upline capture images of Dustin Johnson's driver swing. Image 1 is at P3, image 2 is at P3.5, image 3 is at P4 and image 4 is at P4.5. I have drawn a red line along the length of his thoracic spine and a blue line down the length of his lumbar spine that shows how he acquires left side-bend in his later backswing as he performs an arch-extension maneuver of his thoracic spine between P3.5 => P4.
Note that I have drawn a short green line opposite his lead shoulder at P4 and a short yellow line opposite his lead shoulder at P4.5 - note how his lead shoulder drops groundwards and how it increases lead side-bend. What causes the lead shoulder to temporarily drop groundwards between P4 -> P4.5?
I suspect that two major biomechanical factors are in play. First of all, DJ is performing a "squat move" involving an increased degree of hip joint flexion that causes the spinal bend inclination to increase. Secondly, DJ is actively pulling his two arms groundwards while he is increasingly flexing his thoracic spine. I suspect that the combination of those two biomechanical events provides a causal explanation for the temporary increase in left side-bend that can be seen between P4 => P5.
Here is an overhead swing video of Dustin Johnson's driver swing action. Consider these overhead capture images of DJ's early downswing action. Image 1 is at P4, image 2 is at P4.5 and image 3 is at P5. Let's presume that DJ has 50 degrees of clockwise pelvic rotation at P4 and 100 degrees of upper thoracic spine rotation at P4. That means that he must have ~50 degrees of upper thoracic spine rotation at P4. Note that DJ has a very active pelvic rotation and that he squares his pelvis by P4.5. Note that his upper back is still angled towards the target to a large degree and that he has probably still maintained ~50 degrees of upper torso rotation. Note that DJ has rotated his pelvis to a slightly open alignment by P5 but the logo on the back of his upper shirt is still facing the target to a large degree. He may actually be increasing his degree of X-factor between P4 => P5 even though his upper thoracic spine rotation measurement does not increase (according to that AMG video). Note that DJ gets more wrinkles/creases in his shirt on the right side of his torso just above the level of his pelvis. Milo Lines used to hypothesize that it was due to going from left side-bend to right side-bend - as he describes it in the following video at www.instagram.com/reel/CnwufDIoUDN/ . However, he now believes that it is due to the combination of increased thoracic spine flexion combined with an intact X-factor measurement. I don't disagree with that opinion, but I would also add two additional facts that i) the trail scapula is being protracted as DJ performs an active trail upper arm adduction maneuver and that ii) the pelvis has now become square at P5 - and that causes the distance between the right shoulder socket and the right side of the pelvic rim to decrease. Here is Dustin Johnson's hand arc path between P4 => P5.5. The red splined path represents his hand arc path. Image 1 is at P4, image 2 is at P5 and image 3 is at P5.5. Note that his hands are no closer to the target at P5.5 than they were at P4, and what makes that biomechanically possible is the fact that he is actively pulling his hands downwards/groundwards while he is significantly limiting the amount of any independent counterclockwise upper torso rotation from happening that is greater than the amount that is due to pelvic rotation. Note that his hand arc path is widest at P5 - due to the fact that he is rotating his pelvis to square by P5 while maintaining an intact X-factor, and while he is simultaneously lowering his lead arm from the 11:30 o'clock position towards the 9 o'clock position. A third factor in play that helps to move his hands from its P4 position on the hand arc path to its P5 position on the hand arc path is the fact that he is flexing his thoracic spine by ~40 degrees between P4 => P5 while his chest is facing away from the target. Jeff.
|
|
|
Post by dubiousgolfer on Apr 1, 2024 7:13:17 GMT -5
Dr Mann Isn't the main reason why DJ's hands are no closer to the target at P5.5 than at P4 due to the lower and upper centres of rotation moving laterally towards the target? Look at the top view images of DJ and it looks like his lower and upper body has moved towards target from P4-P5. DG
|
|
|
Post by imperfectgolfer on Apr 1, 2024 9:17:57 GMT -5
Dr Mann Isn't the main reason why DJ's hands are no closer to the target at P5.5 than at P4 due to the lower and upper centres of rotation moving laterally towards the target? Look at the top view images of DJ and it looks like his lower and upper body has moved towards target from P4-P5. DG I think that any motion of the body towards the target should increase the likelihood of the hands becoming closer to the target at P5.5 compared to P4 if the hands were simply moving vertically downwards or if a significant amount of lead arm abduction happened between P4 => P5.5. I think that the major reason why they are not closer is the fact that the trail forearm is angled away from the target so that DJ's hands are >12" outside the right side of the body at P5.5, and also due to the fact that the trail shoulder is further away from the target at P5.5 compared to P4.
Image 1 is at P4, image 2 is at P5 and image 3 is at P5.5.
Jeff.
|
|
|
Post by dubiousgolfer on Apr 1, 2024 11:48:11 GMT -5
Dr Mann,
I think I finally understand what you are saying.
If from P4-P5.5 there were no other biomechanical actions other than:
a. Rotation of pelvis/chest. b. Lateral targetward motion of pelvis/torso. c. Some coupled vertical adduction of trail arm and horizonal abduction of lead arm.
Then DJ's hands would be closer to the target at P5.5 versus P4.
For the hands to not be closer at P5.5 versus their position at P4, would require extra biomechanical actions such as:
d. Protraction of trail scapula. e. Increased thoracic flexion while the chest is closed to the target during top view images 1, 2 and 3.
Have I got this correct?
DG
|
|