Tweet
A number of opinions on the above topics has been expressed on the BIOMCH-L
discussion forum by several colleagues during the last few weeks. It is now
time to present a combined summary for both subject areas. This summary will
be a rather condensed one because I have provided three extensive
intermediary resumes already in my respective postings of 14 January, 15
January, and 25 January.
Apart from a few rather conservative opinions (Dr. G. Duda in his posting of
10 January, for instance, does not believe that "better or more accurate
(biomechanical) models" are required in future developments), the vast
majority of discussants expressed progressive views, rejecting scientific
stagnation in favor of a new impetus for biomechanical research and the
development of improved biomechanical methods. The following PROBLEM AREAS
relating to myoskeletal inverse dynamics, motion analysis, and modeling were
identified as important issues requiring attention and solutions. (The
abbreviation IDMPx will be used to designate "Inverse Dynamics, Modeling,
and Motion Analysis Problem Area no.x").
PROBLEM AREA IDMP1: HUMAN BODY (ANTHROPOMORPHIC) MODELS INCLUDING TISSUE
ANALOGS.
As regards the BODY TISSUE COMPOSITION AND IT'S MASS DISTRIBUTION, the bulk
of the mass is not in the bony structures but in viscoelastic soft tissue
which moves relative to the skeleton. Neglect of these facts leads to
inverse dynamics solutions that are erroneous to varying degrees. The
detailed three-dimensional structures of TERMINAL SEGMENTS (HANDS, FEET) are
not adequately accounted for in present models. Current MODELS OF BODY
JOINTS need to be improved and adequate methods need to be devised for
DETERMINING IN VIVO LOCATIONS AND ORIENTATIONS OF INSTANTANEOUS JOINT AXES
OF ROTATION. Anthropomorphic models must contain a SUFFICIENT NUMBER OF
PARAMETERS that permit adaptation of these models to pathological conditions
on the one hand, and MODEL INDIVIDUALIZATION (personalization) on the other,
including effects of biological adaptation such as remodeling and growth.
Because of the transient nature of the parameter values and the necessity of
biological realism, these values must be determined by appropriately
designed IN VIVO tests. MECHANOMATHEMATICAL MODELS OF THE HUMAN MYOSKELETAL
SYSTEM should be biologically realistic and exist in both, forward and
inverse dynamics formulations. To be useful for clinical purposes they
should include neuromuscular controls.
PROBLEM AREA IDMP2: MOTION DATA RECORDING, CONDITIONING, FILTERING, AND
PROCESSING, INCLUDING TIME DERIVATIVE COMPUTATIONS.
An important and unresolved issue is the SHIFT OF SKIN-ATTACHED MARKERS
RELATIVE TO THE SKELETON, leading to severe distortions of the actual motion
patterns. Closely connected to this problem is that of a more complex body
model in which the behavior of viscoelastic tissue components as well as
that of the skeleton are represented as separate entities. Separate sensors
for skeleton and soft tissue structures would then be attached to the
subject (patient) to record the respective motions. Current methods for
determining OPTIMAL FILTERING WINDOWS for computing numerically zeroth,
first, and second time derivatives of the noisy motion data are inadequate
and need to be improved. Alternatively, well known but more elaborate
techniques such as simulation should also be exploited as possible
substitutes for the ill-posed problem of numerical differentiation.
PROBLEM AREAS TO BE DEFINED LATER:
Some important issues not directly related to myoskeletal inverse dynamics
and motion analysis have been debated by some discussants. In particular,
the identification of neuromyoskeletal performance criteria under
pathological conditions in connection with specific clinical treatment
optimization, outcome prediction, and simulation of surgical options have
been mentioned. These are issues of cardinal practical importance and will
be discussed under BIONET-Topics still to be posted.
I will have to say more about the problem areas listed above in a posting
outside the BIONET context.
Thanks again to all who participated in the discussion of TOPICS 2 and 3.
TOPIC 4 will soon be posted by Dr. Leardini.
For those interested in attending the BIONET EVENT in Brussels, please note
that the CALL FOR POSTERS (http://www.mk.dmu.ac.uk/bionet/event posters.htm)
has now been launched and that REGISTRATION for this event is also open
(http://www.mk.dmu.ac.uk/bionet/event registration.htm). "Early bird"
registration is offered at a discount rate and especially low rates are
available for students and paticipants from Central and Eastern Europe.
Herbert Hatze
************************************************** ********************
Prof. Dr. Herbert Hatze
Head, Department and Laboratory of Biomechanics, ISW,
University of Vienna
Auf der Schmelz 6 Tel: + 43 1 4277 48880
A-1150 WIEN Fax: + 43 1 4277 48889
AUSTRIA e-mail: herbert.hatze@univie.ac.at
************************************************** ********************
---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------
discussion forum by several colleagues during the last few weeks. It is now
time to present a combined summary for both subject areas. This summary will
be a rather condensed one because I have provided three extensive
intermediary resumes already in my respective postings of 14 January, 15
January, and 25 January.
Apart from a few rather conservative opinions (Dr. G. Duda in his posting of
10 January, for instance, does not believe that "better or more accurate
(biomechanical) models" are required in future developments), the vast
majority of discussants expressed progressive views, rejecting scientific
stagnation in favor of a new impetus for biomechanical research and the
development of improved biomechanical methods. The following PROBLEM AREAS
relating to myoskeletal inverse dynamics, motion analysis, and modeling were
identified as important issues requiring attention and solutions. (The
abbreviation IDMPx will be used to designate "Inverse Dynamics, Modeling,
and Motion Analysis Problem Area no.x").
PROBLEM AREA IDMP1: HUMAN BODY (ANTHROPOMORPHIC) MODELS INCLUDING TISSUE
ANALOGS.
As regards the BODY TISSUE COMPOSITION AND IT'S MASS DISTRIBUTION, the bulk
of the mass is not in the bony structures but in viscoelastic soft tissue
which moves relative to the skeleton. Neglect of these facts leads to
inverse dynamics solutions that are erroneous to varying degrees. The
detailed three-dimensional structures of TERMINAL SEGMENTS (HANDS, FEET) are
not adequately accounted for in present models. Current MODELS OF BODY
JOINTS need to be improved and adequate methods need to be devised for
DETERMINING IN VIVO LOCATIONS AND ORIENTATIONS OF INSTANTANEOUS JOINT AXES
OF ROTATION. Anthropomorphic models must contain a SUFFICIENT NUMBER OF
PARAMETERS that permit adaptation of these models to pathological conditions
on the one hand, and MODEL INDIVIDUALIZATION (personalization) on the other,
including effects of biological adaptation such as remodeling and growth.
Because of the transient nature of the parameter values and the necessity of
biological realism, these values must be determined by appropriately
designed IN VIVO tests. MECHANOMATHEMATICAL MODELS OF THE HUMAN MYOSKELETAL
SYSTEM should be biologically realistic and exist in both, forward and
inverse dynamics formulations. To be useful for clinical purposes they
should include neuromuscular controls.
PROBLEM AREA IDMP2: MOTION DATA RECORDING, CONDITIONING, FILTERING, AND
PROCESSING, INCLUDING TIME DERIVATIVE COMPUTATIONS.
An important and unresolved issue is the SHIFT OF SKIN-ATTACHED MARKERS
RELATIVE TO THE SKELETON, leading to severe distortions of the actual motion
patterns. Closely connected to this problem is that of a more complex body
model in which the behavior of viscoelastic tissue components as well as
that of the skeleton are represented as separate entities. Separate sensors
for skeleton and soft tissue structures would then be attached to the
subject (patient) to record the respective motions. Current methods for
determining OPTIMAL FILTERING WINDOWS for computing numerically zeroth,
first, and second time derivatives of the noisy motion data are inadequate
and need to be improved. Alternatively, well known but more elaborate
techniques such as simulation should also be exploited as possible
substitutes for the ill-posed problem of numerical differentiation.
PROBLEM AREAS TO BE DEFINED LATER:
Some important issues not directly related to myoskeletal inverse dynamics
and motion analysis have been debated by some discussants. In particular,
the identification of neuromyoskeletal performance criteria under
pathological conditions in connection with specific clinical treatment
optimization, outcome prediction, and simulation of surgical options have
been mentioned. These are issues of cardinal practical importance and will
be discussed under BIONET-Topics still to be posted.
I will have to say more about the problem areas listed above in a posting
outside the BIONET context.
Thanks again to all who participated in the discussion of TOPICS 2 and 3.
TOPIC 4 will soon be posted by Dr. Leardini.
For those interested in attending the BIONET EVENT in Brussels, please note
that the CALL FOR POSTERS (http://www.mk.dmu.ac.uk/bionet/event posters.htm)
has now been launched and that REGISTRATION for this event is also open
(http://www.mk.dmu.ac.uk/bionet/event registration.htm). "Early bird"
registration is offered at a discount rate and especially low rates are
available for students and paticipants from Central and Eastern Europe.
Herbert Hatze
************************************************** ********************
Prof. Dr. Herbert Hatze
Head, Department and Laboratory of Biomechanics, ISW,
University of Vienna
Auf der Schmelz 6 Tel: + 43 1 4277 48880
A-1150 WIEN Fax: + 43 1 4277 48889
AUSTRIA e-mail: herbert.hatze@univie.ac.at
************************************************** ********************
---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------