Overview of morphological parameters for modelling of the human shoulder

The VU-study (1988-1996)
 

H.E.J. Veeger

Vrije Universiteit Amsterdam, Delft University of Technology

F.C.T. van der Helm

Delft University of Technology

 

(to overview of all data)
 

These data have been made available for non-commercial use only. In any other case, previous contact with the authors is necessary. Any use of these data should be accompanied with proper reference to the source. It is suggested to use the following references when referring to these data:

Veeger, H.E.J., Helm, F.C.T., Woude, L.H.V. van der, Pronk, G.M. and Rozendal, R.H. (1991) Inertia and muscle contraction parameters for musculoskeletal modelling of the shoulder mechanism. Journal of Biomechanics. 24, 615-631.

 

Helm, F.C.T. van der, Veeger, H.E.J., Pronk, G.M., Woude, L.H.V. van der and Rozendal, R.H. (1992) Geometry parameters of musculoskeletal modelling of the shoulder mechanism. J. Biomech. 25, 129-144.


Preface

These pages contain (most of the) anthropometric data that have been collected in a dissection study focussing on the human shoulder. This study was performed between 1988 and 1990, at the Vrije Universiteit Amsterdam. Its purpose was to collect data for musculoskeletal modelling of the human shoulder. We dissected seven specimen, from which we took both left and right shoulders. For each specimen, anthropometric data were collected before dissection. Subsequently, all relevant muscle attachments and bony elements on the thorax, scapula, clavicle and humerus were digitized. The dissection procedure ended with the measurement of the Physiological Cross-sectional Area of each muscle.

Given are the anthropometric data for all seven specimen and the attachments for all right shoulders (N=7). The attachment data are given in different formats. We tried to make these data as easy to read as possible. To facilitate interpretation, we added a brief description of the data collection and processing procedures.

DirkJan Veeger


Table of contents (this document)

    1. Introduction
    2. Description of origins and insertions of muscles
    3. Reconstruction of muscle geometry
    4. Description of bony elements and articular surfaces

 

 

 


 

Available data (tables)

Inertial data

Anatomical landmarks

(both left and right shoulders)

collum humeri modeled as a cylinder

 

Proportional elements

·  NL-K1_R

·  NL_K2_R

·  NL_K3_R

·  NL_K4_R

·  NL_K5_R

·  NL_K6_R

·  NL_K7_R

Origins & Insertions expressed as t-values

·  NL-K1_R

·  NL_K2_R

·  NL_K3_R

·  NL_K4_R

·  NL_K5_R

·  NL_K6_R

·  NL_K7_R

Mapping of Origins to Insertions

·  NL-K1_R

·  NL_K2_R

·  NL_K3_R

·  NL_K4_R

·  NL_K5_R

·  NL_K6_R

·  NL_K7_R



 
 

Introduction

To develop a musculoskeletal model of the shoulder mechanism, both shoulders of seven cadavers were measured to obtain a complete set of parameters.
Using antropometric measurements, the mass and rotational inertia of segments were estimated, followed by three-dimensional measurements of all morphological structures relevant for modelling, i.e. muscle origins and insertions, muscle bundle directions, ligament attachments and articular surfaces; all in relation to selected bony landmarks. Subsequently, muscle contraction parameters as muscle mass and physiological cross-sectional area were measured.
The method of data collection and the results for inertia and muscle contraction parameters as prerequisities for modelling are described. In addition, a dynamical finite-element model of the shoulder mechanism consisting of thorax, clavicula, scapula and humerus is outlined.
Also, the geometry parameters that form the basis of the musculoskeletal muscle are described. Morphological structures are modelled as geometrical forms. Parameters describing this form are estimated from 3-D position coordinates of a large number of datapoints on the morphological structure, using a least-squares criterion.
Muscle and ligament attachments are represented as a plane or as a (curved) line. Muscle paths are determined by a geometrical form of the bony contour around which the muscle is wrapped. Muscle architecture is determined by the distribution of muscle bundles over the attachment area, mapping the distribution of the origin to the insertion. Joint rotation centers are derived from articular surfaces. Hence, muscle moment arms can be calculated. The result of this study is a set of parameters for each cadaver, describing very precisely the geometry of the shoulder mechanism. This set allows positioning of muscle force vectors a posteriori, and the recalculation of position coordinates and moment arms for any position of the shoulder.
 

Methods

During the measurements, the following sequence of steps was made:

1. measurement of relevant body dimensions for the derivation of inertia parameters segment mass, segment volume, segment mass position and moments of inertia.
Body dimensions were measured following the procedures and definitions as given by Clauser et al (1969) and used for the estimation of segment properties on the basis of regression equations published by those same authors and by Hinrichs (1985).

2. measurement of well-palpable bony landmarks for future comparison with in-vivo measurements.
Anatomical landmarks were measured with a 3-D digitizer (Figure1). The landmarks were measured with the specimen still intact and together with the screws that were used to define technical coordinate systems (see 3.).
The 3-D data collection was performed with a spatial digitizer, the "palpator" (See Pronk & van der Helm, 1991). The palpator consists of an open chain of four links of 0.2 m each, connected by four perpendicular hinges. The rotation of each hinge is recorded with a potentiometer and is on-line AD-converted (Olivetti M21, DT2801 AD converter). The 3D position of the end point of the fourth link can be calculated with respect to an internal coördinate system of the palpator, defined previously in calibration measurements. Measurement error was estimated to have an standard deviation of 0.96 mm per coördinate or 1.43 mm in absolute distance (Pronk & van der Helm, 1991)

3. measurements of small screws in each segment, for construction of technical coordinate systems, with the specimen fully intact.
In each segment at least four srewes were fastened. All screws were measured in the intact, upright specimen (see Figure 2). These screws were subsequently used for definition of the technical coordinate system and for the definition of the global position of the specimen.
 

Image4.gif
 

Figure 1

The palpator


Image3.gif

Figure 2

Illustration of the measurement rig during measurements aimed at defining the global orientation of local technical coordinate systems.

4. dissection and exarticulation of each specimen at the scapulothoracic, acromioclavicular, sternoclavicular and glenohumeral joints.

5. measurement of origins and insertions of muscles, shapes of articular surfaces and shapes of relevant (parts of) bony elements.
With the use of the palpator, all origins and insertions, articular surfaces and shapes of potential 'muscle wrapping' areas were digitized, always relative to their own technical coordinate system. The technical coordinate system was defined by the same screws that were used in 3). The number of digitized points was variable, and mostly based on visual judgment whether the area to be measured was sufficiently described. This implied that for muscles with point-shaped origins at least five points were taken, but that the number of points could also be as high as 100 for muscles with large attachment areas such as M. subscapularis. With the use of the technical coordinate systems, all local coordinates were transformed to the global coordinate system as desribed previously. Subsequently for each cadaver a global coordinate system was defined related to the anatomical conventions as a system with its origin in the Incisura Jugularis, the Y-axis directed cranially, the Z-axis in the sagittal plane towards the back and the X-axis aling the frontal plane towrds the right shoulder.

6. measurements of muscle mass, volume and PCSA
Physiological Cross Section Area (PCSA) and muscle-mass were measured on both left and right side of the body (Veeger et al, 1991).
The PCSA was measured directly. Of each muscle a cross-section was made by cutting the muscle at its bulkiest part, perpendicular to the muscle bundles. For the 'flat' muscles, the resulting strips of muscle were rolled up. All cuts were photographed and the projected surfaces were later digitized. Statistical comparison on the existence of left-right differences for PCSA or muscle-mass within each body (T-test, 16 or 17 muscles, p < 0.01) did not lead to significant results.
 

Description of origins and insertions of muscles

The attachment sites of the muscles (Origin and insertion) are described as points, lines, or planes, where a point is a zero-order polynome. All data are given in the position in which the cadaver was measured, i.e. not the anatomical position. However, since the positions of the bones are also given, any chosen position can be obtained.
For those attachment that can be described as points or lines, the attachments are parameterized as a function of  t. Using vectors t x, t y , t z an attachment line (or a point as a zero-order polynome) can be described as a function of t (0 " t " 1):

x = t x (1) + t x (2).t + t x (3).t 2 + ..... + t x (n).t n-1

y = t y (1) + t y (2).t + t y (3).t 2 + ..... + t y (n).t n-1

z = t z (1) + t z (2).t + t z (3).t 2 + ..... + t z (n).t n-1

where t x (1) is the first element of vector t x , etc. Start and end of the attachment line is given by t = 0 and t = 1 , respectively. In the sheets "a" stands for "X", "b" for "Y" and "c" for "Z" (Dimension is in mm). See for the results on this  the tables with Origins & Insertions expressed as t-values.

Reconstruction of muscle geometry

To be able to reconstruct this muscle architecture, colored beads were used to mark origin and insertion of some muscle bundles. For a number of corresponding beads x-, y-, and z-coördinates of origin and insertion were recorded. After a line was fitted to the attachment, a t-value has been assigned to each point at the attachment which was marked with a bead. Hence, for N beads t-values at the origin t oj (j= 1, ... N), and t-values at the insertion t ij are known. Next, t oj is fitted to t ij. This map can be used to interpolate corresponding t-values of all fibers in between. Using these maps in combination with the geometry parameters, the attachment site of any muscle bundle can be reconstructed (see Figures 3 and 4). See Mapping of Origins to Insertions for the results of this procedure.
The distribution map is also used to locate a large number of elements (± 200), which will account for the influence of muscle geometry on resulting force and moments of the muscle. However, this would require calculation of the effect of those 200 elements in each position of the mechanism which is inefficient in terms of computing time. Therefore the number of elements is minimised to the minimum of force vectors necessary to represent the muscle (van der Helm & Veenbaas, 1991)
Figure 3 shows the number of degrees of freedom which can be influenced (=number of independent force vectors) for several combinations of attachment sites. Lines of second or higher order can be regarded similarly as planes: they provide three non-collinear points.
 

Image6.gif

Image7.gif

Figure 3
Number of elements needed to define a muscle that can be described as having point-shaped, line-shaped or plane-shaped attachments.

Figure 4
The lateral rotating torque of m.trapezius. When represented with one force vector the torque can never be modelled; with two or more force vectors the distribution over the attachment site and direction of the force vector is critical for muscle force calculation

Description of bony elements and articular surfaces

Bony structures like joint surfaces and relevant attachment sites can were described mathematically by geometric forms (points, lines, cylinders, spheres, ellipsoids) which are fitted to the data points using a least squares criterion (van der Helm et al, 1992). Flat attachment sites for instance have been modeled as planes. Also, the humeral head has been modeled as a sphere, whereas the the humeral shaft has been modeled as a cilinder. A full description of the parameter estimations has been given in the appendixes of the 1992 paper (van der Helm et al, 1992). Not all of these data are yet (september 17, 1999) available on these net pages.
 

References

Clauser CE, McConville JT, Young JM (1969) Weight, volume and center of mass of segments of the human body. AMRL-TR-69-70, Wright Patterson Airforce Base, Dayton, Ohio.
Hinrichs RN (1985) Regression equations to predict segmental moment equations fro anthropometric measurements: an extension of the data of Chandler et al (1975). J. Biomech.18, 621-624.
Pronk GM, Helm, F.C.T. van der (1991) The palpator, an instrument for measuring the three-dimensional positions of bony landmarks in a  fast and easy way. J. Med. Eng. Techn. 15(1), 15-20.
Helm, F.C.T. van der , Veenbaas R. (1991) Modelling the mechanical effect of muscles with large attachment sites: an application to the shoulder mechanism. J. Biomech. 24,
Veeger, H.E.J., Helm, F.C.T., Woude, L.H.V. van der, Pronk, G.M. and Rozendal, R.H. (1991) Inertia and muscle contraction parameters for musculoskeletal modelling of the shoulder mechanism. Journal of Biomechanics. 24, 615-631.
Helm, F.C.T. van der, Veeger, H.E.J., Pronk, G.M., Woude, L.H.V. van der and Rozendal, R.H. (1992) Geometry parameters of musculoskeletal modelling of the shoulder mechanism. J. Biomech. 25, 129-144.