How many markers needed for cervical spine?

Hi, AnyBody Support,

I am cathy. My research effort is focused on the cervical spine part. This time I need to determine how many markers that I need to put on the head/neck to run the model, not over- or under-constrained. Then I will do the experiment and get the c3d data.

My calculations are shown as below:

9 segments: skull, c1-c7, T1: (Total dof: 6 * 9 = 54 ).

8 joints: 2 revolute joints (c0c1, c1c2), so 2 constraints here;
6 spherical joints (c2c3, c3c4, c4c5, c5c6, c6c7, c7T1), so
6 * 3 = 18 constraints here.

Then, 54 -2 -18 = 34 marker coordinates are required from experiment. Correct?

One problem here: if I pick x, y, z coordinates from each marker, then 12 markers will be enough to provide 34 marker coordinates? or 34 markers needed?

Finally, to put markers on the body:
Head: 3 markers; (c0c1 joint is inside the skull)
C2 - C7: 6 * 3 = 18 markers; (Is it too much to put 3 tiny markers on each segment?)
T1: 3 markers.

-----------I will put 24 markers in total.
Do I still/really need 10 more markers to meet my goal?----------

[b]Is my calculation correct? If incorrect, please tell me where is wrong.

How many markers are needed to make the model run smoothly?[/b]

Sorry for my bothering you, since I do not have any mechanical engineering background.

Thank you! Your help is really appreciated!


Hi Cathy,
each marker will give you 3 contraints (x,y,z). AnyBody in the current version can handle overdetermined systems, so basically the equation is the more the better. The cervical spine is equipped with a spine rhythm, similar as the lumbar spine. Therefore, the open degrees of freedom are only between skull and thorax. Please have a look at the JointTrunkLumbarNeck files. If you make use of this only a few markers on the skull and the thorax are enough. If you want to drive each vertebral body with real coordinates you have to follow an approach as you have described.
Best regards,

Hi, Sebastian,

Thanks a lot. Your answer helped me a lot.
But I don’t understand this: “Therefore, the open degrees of freedom are only between skull and thorax.” What does this mean? Can you please explain it in more detail?

I am thinking one method using Kinematic Measures:

Because it will not be easy to put that many markers on the exact cervical segment precisely.

If my c3d file does not have enough markers (on the neck) to drive the cervical vertebrae very well, can I use the kinematic measures method and let the cervical vertebrae follow the motion of head? or the shoulder’s motion? Which one is correct? Is it still possible to get the correct output of neck muscles’ activity?

Thanks for your patience!

Enjoy the summer!


Hi Cathy,

I think the cervical rhythm Sebastian told you about is already doing what you call the “kinematic measure method”.
In this rhythm the motion of the cervical joints are linked together, like C7-C6 flexion equal to C6-C5 flexion and so on. So you just have to drive one global flexion angle between thorax and skull and all the cervical vertebra will automatically flex in between accoding to this rhythm.

I hope this is clear. Same principle for lateral bending and axial rotation.

Best regards, Sylvain.