Thank you all who responded to my question about inverse dynamic 'golden

standard'. Original question is followed by five replays.

Regards, Tomislav.

QUESTION:

Dear All,

concerning recent discussion about inverse dynamic I would like to ask the

following. How do we actually validate accuracy of kinetic parameters

(force, moments, power etc.) obtained through the inverse dynamic

calculation? What is used as a 'golden standard' ? Namely, inverse dynamic

output basically relay on two input sources.

Firstly, kinematic parameters acquired usually by the means of 3D kinematic

reconstruction systems. I have seen many papers arguing the accuracy of the

reconstructions systems themselves and some papers discussing the accuracy

of joint center determination against some golden standard (e.g. 'roentgen

stereophotogrammetry). Here, the hip arises as significant challenge and

generally skin effect imposed on marker setup.

Second input comes from so called BSP parameters calculation (m, I, CM, V,

r). In this cases people again tested their body models and corresponding

BSPs against some golden standards (e.g. CT, MRI, gamma mass scanning, DXA

(dual energy X ray absompthometry), pendulum method etc.).

However, is there any way to validate the actual numbers for force, moments

and/or power exercised at specific joint using some other golden standard?

Or the best we can do is to relay on accuracy estimation for 3D kinematic

systems and BSP, and then analytically predict how the final inverse dynamic

calculations accurate will be (including perhaps in our estimation other

inherent assumption about inverse dynamic)?

Thank you for your response. Summary will follow later on.

Regards, Tomislav.

Tomislav Pribanic, M.Sc., EE

Department for Electronic Systems and Information Processing

Faculty of Electrical Engineering and Computing

3 Unska, 10000 Zagreb, Croatia

tel. ..385 1 612 98 67, fax. ..385 1 612 96 52

E-mail : tomislav.pribanic@fer.hr

REPLAY1:

Some references on model validation.

Regards,

Angela J. Tate, M.Sc.

Doctoral Candidate

Faculty of Engineering and Applied Science

Memorial University of Newfoundland

St. John's, NL CANADA A1B 3X5

tel. (709) 737-8994

fax. (709) 737-4042

angelat@engr.mun.ca

www.engr.mun.ca\~angelat

De Looze, M.P., Kingma, I., Bussmann, J.B.J., & Toussaint, H.M., 1992.

Validation of a dynamic linked segment model to calculate joint moments in

lifting, Clin. Biomech. 7, 161-169.

Desjardins, P., Plamondon, A., & Gagnon, M., 1998. Sensitivity analysis of

segment models to estimate the net reaction moments at the L5/S1 joint in

lifting, Med. Eng. Phys. 20(2), 153-158.

Frigo, C., 1990. Three-dimensional model for studying the dynamics loads on

the spine during lifting., Clin. Biomech. 5, 143-152.

Kingma, I., De Looze, M.P., Toussaint, H.M., Klijnsma, H.G., & Bruijnen,

B.M., 1996a. Validation of a full body 3-D dynamic linked segment model,

Hum. Movement Sci., 15, 833-860.

Kingma, I., Toussaint, H.M., De Looze, M.P., Van Dieen, J.P., 1996b. Segment

inertial parameter evaluation in two anthropometric models by application of

a dynamic linked segment model. J. Biomech., 29(5), 693-704.

Lariviere, C. & Gagnon, D., 1998. Comparison between two dynamic methods to

estimate triaxial net reaction moments at the L5/S1 joint during lifting,

Clin. Biomech., 13(1), 36-47.

Lariviere, C. & Gagnon, D., 1999a. The influence of trunk modelling in 3D

biomechanical analysis of simple and complex lifting tasks, Clin. Biomech.

14(7), 449-461.

Lariviere, C. & Gagnon, D., 1999b. The L5/S1 joint moment sensitvity to

measurement errors in dynamic 3D multisegment lifting models. Hum. Movement

Sci. 18, 573-587.

Plamondon, A., Gagnon, M., & Desjardins, P., 1996. Validation of two 3-D

segment models to calculate the net reaction forces and moments at the

L(5)/S(1) joint in lifting, Clin. Biomech. 11(2), 101-110.

REPLAY2:

----- Original Message -----

From: D. Gordon E. Robertson, Ph.D.

To: Tomislav Pribanic

Sent: Tuesday, February 10, 2004 5:09 PM

Subject: Re: [BIOMCH-L] inverse dynamic calcuation validation

One way that I used is to compare the rate of change of energy of individual

segments with the power supplied to the segment by the forces and moments of

force at the segment's connection with the rest of the body or the

environment. This is sometimes called the "energy" or "power balance." The

idea is that the forces and moments are computed using inverse dynamics and

are therefore affected by the segment's BSPs, kinematics and external

forces, such as the ground reaction force. The changes in mechanical energy

do not involve inverse dynamics, but only on the accuracy of the kinematics

and BSPs. Have a look at the paper:

Robertson, D.G.E. and Winter, D.A. (1980) Mechanical energy generation,

absorption and transfer amongst segments during walking. Journal of

Biomechanics, 13: 845-854.

see also

Siegel, K.L.; Kepple, T.M. & Caldwell, G.E. (1996) Improved agreement of

foot segmental power and rate of energy during gait: Inclusion of distal

power terms and use of 3D models. Journal of Biomechanics, 29:823827

REPLAY3:

Dear Tomislav:

One of the simplest ways to check the accuracy of at least some measurements

and assumptions is to compute the acceleration of the GCM of the entire body

during a free flight (e.g. in running or jumping). Ideally it should be

close to g = -9.8 m/s^2.

Another option is to compare the actual and computed GRF data for a one-leg

support period in walking or running.

The accuracy of the BSP data can be estimated by comparing predicted and

measured GCM position using a reaction board.

Sincerely,

Vladimir Zatsiorsky

REPLAY4 (articles available on request since list doe not support

attachment):

----- Original Message -----

From: "Tim L. A. Doyle"

To: "'Tomislav Pribanic'"

Sent: Wednesday, February 11, 2004 1:15 AM

Subject: RE: [BIOMCH-L] inverse dynamic calcuation validation

You might find these articles useful.

Tim.

REPLAY5:

> dear Tomislav

>

> I had the same problem for my PhD project for the calculation of inverse

> dynamics starting from upper limbs kinematics

> I'm sorry to say that I didn't find any "golden standard" neither for the

> accuracy of 3D reconstruction nor for BSP parameters

> I tried to evaluate the accuracy of 3D reconstruction basing on pc

> simulations, giving to my model as input first, some "virtual and not

noisy"

> coordinates and then some "real and noisy" coordinates, then calculating

the

> differences to have an idea of the errors I was including due to 3D

> reconstruction.

> As for BSP, I think that was the major source of errors since I had to

take

> data from the literature and the number of subjects investigated was

always

> very low...

> I think this topic is very important -probably it will generate a tough

> discussion- and I hope something good as a golden standard will arise from

> the list

> Greetings from Italy

>

> Federica

> %-------------------------------------------------%

> Federica Sibella, PhD

> Projects&Technology Department

> viale Marconi 5, 24044 Dalmine (Bergamo), Italy

> tel: +39 035 205 2361

> e-mail: federica.sibella@unibg.it

-----------------------------------------------------------------

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl

For information and archives: http://isb.ri.ccf.org/biomch-l

Please consider posting your message to the Biomch-L Web-based

Discussion Forum: http://movement-analysis.com/biomch_l

-----------------------------------------------------------------

standard'. Original question is followed by five replays.

Regards, Tomislav.

QUESTION:

Dear All,

concerning recent discussion about inverse dynamic I would like to ask the

following. How do we actually validate accuracy of kinetic parameters

(force, moments, power etc.) obtained through the inverse dynamic

calculation? What is used as a 'golden standard' ? Namely, inverse dynamic

output basically relay on two input sources.

Firstly, kinematic parameters acquired usually by the means of 3D kinematic

reconstruction systems. I have seen many papers arguing the accuracy of the

reconstructions systems themselves and some papers discussing the accuracy

of joint center determination against some golden standard (e.g. 'roentgen

stereophotogrammetry). Here, the hip arises as significant challenge and

generally skin effect imposed on marker setup.

Second input comes from so called BSP parameters calculation (m, I, CM, V,

r). In this cases people again tested their body models and corresponding

BSPs against some golden standards (e.g. CT, MRI, gamma mass scanning, DXA

(dual energy X ray absompthometry), pendulum method etc.).

However, is there any way to validate the actual numbers for force, moments

and/or power exercised at specific joint using some other golden standard?

Or the best we can do is to relay on accuracy estimation for 3D kinematic

systems and BSP, and then analytically predict how the final inverse dynamic

calculations accurate will be (including perhaps in our estimation other

inherent assumption about inverse dynamic)?

Thank you for your response. Summary will follow later on.

Regards, Tomislav.

Tomislav Pribanic, M.Sc., EE

Department for Electronic Systems and Information Processing

Faculty of Electrical Engineering and Computing

3 Unska, 10000 Zagreb, Croatia

tel. ..385 1 612 98 67, fax. ..385 1 612 96 52

E-mail : tomislav.pribanic@fer.hr

REPLAY1:

Some references on model validation.

Regards,

Angela J. Tate, M.Sc.

Doctoral Candidate

Faculty of Engineering and Applied Science

Memorial University of Newfoundland

St. John's, NL CANADA A1B 3X5

tel. (709) 737-8994

fax. (709) 737-4042

angelat@engr.mun.ca

www.engr.mun.ca\~angelat

De Looze, M.P., Kingma, I., Bussmann, J.B.J., & Toussaint, H.M., 1992.

Validation of a dynamic linked segment model to calculate joint moments in

lifting, Clin. Biomech. 7, 161-169.

Desjardins, P., Plamondon, A., & Gagnon, M., 1998. Sensitivity analysis of

segment models to estimate the net reaction moments at the L5/S1 joint in

lifting, Med. Eng. Phys. 20(2), 153-158.

Frigo, C., 1990. Three-dimensional model for studying the dynamics loads on

the spine during lifting., Clin. Biomech. 5, 143-152.

Kingma, I., De Looze, M.P., Toussaint, H.M., Klijnsma, H.G., & Bruijnen,

B.M., 1996a. Validation of a full body 3-D dynamic linked segment model,

Hum. Movement Sci., 15, 833-860.

Kingma, I., Toussaint, H.M., De Looze, M.P., Van Dieen, J.P., 1996b. Segment

inertial parameter evaluation in two anthropometric models by application of

a dynamic linked segment model. J. Biomech., 29(5), 693-704.

Lariviere, C. & Gagnon, D., 1998. Comparison between two dynamic methods to

estimate triaxial net reaction moments at the L5/S1 joint during lifting,

Clin. Biomech., 13(1), 36-47.

Lariviere, C. & Gagnon, D., 1999a. The influence of trunk modelling in 3D

biomechanical analysis of simple and complex lifting tasks, Clin. Biomech.

14(7), 449-461.

Lariviere, C. & Gagnon, D., 1999b. The L5/S1 joint moment sensitvity to

measurement errors in dynamic 3D multisegment lifting models. Hum. Movement

Sci. 18, 573-587.

Plamondon, A., Gagnon, M., & Desjardins, P., 1996. Validation of two 3-D

segment models to calculate the net reaction forces and moments at the

L(5)/S(1) joint in lifting, Clin. Biomech. 11(2), 101-110.

REPLAY2:

----- Original Message -----

From: D. Gordon E. Robertson, Ph.D.

To: Tomislav Pribanic

Sent: Tuesday, February 10, 2004 5:09 PM

Subject: Re: [BIOMCH-L] inverse dynamic calcuation validation

One way that I used is to compare the rate of change of energy of individual

segments with the power supplied to the segment by the forces and moments of

force at the segment's connection with the rest of the body or the

environment. This is sometimes called the "energy" or "power balance." The

idea is that the forces and moments are computed using inverse dynamics and

are therefore affected by the segment's BSPs, kinematics and external

forces, such as the ground reaction force. The changes in mechanical energy

do not involve inverse dynamics, but only on the accuracy of the kinematics

and BSPs. Have a look at the paper:

Robertson, D.G.E. and Winter, D.A. (1980) Mechanical energy generation,

absorption and transfer amongst segments during walking. Journal of

Biomechanics, 13: 845-854.

see also

Siegel, K.L.; Kepple, T.M. & Caldwell, G.E. (1996) Improved agreement of

foot segmental power and rate of energy during gait: Inclusion of distal

power terms and use of 3D models. Journal of Biomechanics, 29:823827

REPLAY3:

Dear Tomislav:

One of the simplest ways to check the accuracy of at least some measurements

and assumptions is to compute the acceleration of the GCM of the entire body

during a free flight (e.g. in running or jumping). Ideally it should be

close to g = -9.8 m/s^2.

Another option is to compare the actual and computed GRF data for a one-leg

support period in walking or running.

The accuracy of the BSP data can be estimated by comparing predicted and

measured GCM position using a reaction board.

Sincerely,

Vladimir Zatsiorsky

REPLAY4 (articles available on request since list doe not support

attachment):

----- Original Message -----

From: "Tim L. A. Doyle"

To: "'Tomislav Pribanic'"

Sent: Wednesday, February 11, 2004 1:15 AM

Subject: RE: [BIOMCH-L] inverse dynamic calcuation validation

You might find these articles useful.

Tim.

REPLAY5:

> dear Tomislav

>

> I had the same problem for my PhD project for the calculation of inverse

> dynamics starting from upper limbs kinematics

> I'm sorry to say that I didn't find any "golden standard" neither for the

> accuracy of 3D reconstruction nor for BSP parameters

> I tried to evaluate the accuracy of 3D reconstruction basing on pc

> simulations, giving to my model as input first, some "virtual and not

noisy"

> coordinates and then some "real and noisy" coordinates, then calculating

the

> differences to have an idea of the errors I was including due to 3D

> reconstruction.

> As for BSP, I think that was the major source of errors since I had to

take

> data from the literature and the number of subjects investigated was

always

> very low...

> I think this topic is very important -probably it will generate a tough

> discussion- and I hope something good as a golden standard will arise from

> the list

> Greetings from Italy

>

> Federica

> %-------------------------------------------------%

> Federica Sibella, PhD

> Projects&Technology Department

> viale Marconi 5, 24044 Dalmine (Bergamo), Italy

> tel: +39 035 205 2361

> e-mail: federica.sibella@unibg.it

-----------------------------------------------------------------

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl

For information and archives: http://isb.ri.ccf.org/biomch-l

Please consider posting your message to the Biomch-L Web-based

Discussion Forum: http://movement-analysis.com/biomch_l

-----------------------------------------------------------------