SEARCH

SEARCH BY CITATION

Keywords:

  • glenohumeral;
  • glenoid;
  • humeral head;
  • measurement;
  • protractor;
  • shoulder

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Author contributions
  8. References

Functional biomechanics studies of the glenohumeral (GH) soft tissues require an understanding of their sites of bony attachment. Anatomical positions of GH capsular structures have often been quantified relative to the rims of the glenoid and humeral head (HH). The aim of this study was twofold: (1) to quantify the reliability of a set of protractors that directly fit on to the glenoid and HH rims and (2) to use this to determine direct angular position referencing of landmarks and soft tissue attachment points. Three assessors independently used the protractors to assess nine prescribed landmarks on 30 dry bone specimens (15 glenoids and 15 HHs) recording the angular positions of the structures relative to the glenoid and HH. The collected data showed high levels of validity as indicated by the protractor’s intra- and inter-assessor reliabilities: 98.2 and 98.7% for the glenoid component, and 96.2 and 96.5% for the humeral component, respectively. The device could be useful in anatomical studies, description of defects and pathologies on glenohumeral articulation, and planning of scapular reconstructive surgery.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Author contributions
  8. References

Precise anatomical descriptions of sites of structures at the glenohumeral joint (GHJ) are important in biomechanics and during some capsular reconstructive surgeries such as those involving anchor/suture insertion (Bankart, 1938; Lehtinen et al. 2004). Some anatomical structures and landmarks around the human glenoid and the humeral head (HH) are often described relative to the face-of-the-clock or 360°-of-the-circle coordinate system (Bigliani et al. 1992; Boardman et al. 1996; Gagey & Gagey, 2000; Sugalski et al. 2005). Although this convention has been used to report the anatomical bony attachment of soft tissues such as ligaments, it remains unclear how these anatomical structures and landmarks were measured and what appropriately designed geometric instrumentation was used. A set of bone-shape tailored protractors has recently been devised to allow direct measurements of anatomical landmarks and structures relative to the rims of the glenoid and the HH in a 360° coordinate system (Fig. 1). The applicability of the devised protractors could only be broadened if these were shown to be accurate and aiding repeatable and reliable quantifications. The aims of this study were to evaluate the reliability of the use of this tool and to compare its quantifications with known values.

image

Figure 1.  Humeral and glenoid protractor components (protractors are graduated increasingly in the anatomical anterior sense).

Download figure to PowerPoint

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Author contributions
  8. References

Thirty dry specimens of the GHJ bones (15 scapulae and 15 humeri) were used for this study. The protractor origin (0° point) on the glenoid and humeral components were fixed at the superior-central point of the supraglenoid tubercle and at the superior-middle midsubstance of the humeral anatomical neck, respectively (Amadi & Bull, 2011). The protractors, graduated 0–360° anteriorly, were manufactured in five different sizes scaled to equal height-width aspect ratios each, covering the literature-published ranges of human HH diameters and glenoid widths and heights (Robertson et al. 2000; Churchill et al. 2001; Amadi et al. 2008, 2009).

Three assessors were selected and instructed on the locations of the protractor origins of the bones. They were detailed to select independently the best fitting protractor for each bone and blindly record the anatomical positions of seven other notable anatomical landmarks (three humeral and four glenoid) as shown in Fig. 2. The choice of appropriate protractor size for a specimen could be guided by the size that could pass through the hemispherical aspect of the HH to fit on the anatomical neck but without being too loose to wobble during measurement. The glenoid component could be selected based on a height and width approximately equal with the specimen’s to reduce estimation errors. The measured landmarks were: the superior-middle midsubstance of the anatomical neck, the humeral fovea capitis, mid-point of the medial aspect of the lesser tuberosity, mid-point of the medial aspect of the surgical neck, superior-middle point of the supraglenoid tubercle, the anterior point of the neck of the glenoid rim, and the most anterior, caudal and dorsal points of the glenoid rim. These well defined anatomical regions on the bones still require expert assessor interpretation. Each assessor made two trial measurements, each separated by a minimum of 1-h recreational break on the same day. The time taken for each round of measurements was also noted. The average time taken to quantify a landmark point with the protractors particularly confirms how easily users can cross-match an assumed bone origin for the protractors and read off a landmark of interest. The average overall location of a landmark as quantified from the three assessors was applied to compare protractor measurement with known values in the literature.

image

Figure 2.  Quantified landmarks: (a) the superior-middle midsubstance of the anatomical neck, (b) the fovea capitis (FC) of the humerus, (c) the mid-point of the medial aspect of the lesser tuberosity, (d) the mid-point of the medial aspect of the surgical neck, (e) the superior-middle point of the supraglenoid tubercle, (f) the anterior point of the neck of the glenoid rim, (g) the most anterior point of the glenoid rim, (h) the most caudal point of the glenoid rim, (i) the most dorsal point of the glenoid rim.

Download figure to PowerPoint

Intra-assessor repeatability for a landmark (x)

This was a measure of how accurately the use of the protractors by an assessor obtained the same value for the same landmark each time. A generalised approach for the random error in repeated trial measurements in the use of the instrument by assessors was considered (Taylor, 2011). The mean T of N trials (T1 to TN) was used to quantify an assessor’s fractional error of measurement for a landmark of a specimen thus: Fractional error = inline image; where inline image is the absolute value of the difference between a repeat trial Ti and the mean Tα. For the present case of N = 2, i.e. (T1 and T2 only), this simplifies to:

Fractional error = inline image

The average percentage error for a landmark over the 15 specimens was quantified as:

  • image

Over the seven landmarks and 15 specimens for one assessor (n), the average error was quantified as:

  • image

Finally, the overall average error across three assessors (= 1–3) = inline image

Percentage accuracy, and hence intra-assessor reliability, was quantified as R = (100 − E)%.

Inter-assessor accuracy for a landmark

This was a measure of how any user could use the protractors to assign the same angular value to the same landmark. The overall average of the repeated trials of a landmark was compared across the three assessors and used to compute the inter-assessor accuracy. Each assessor’s average measurement (M1, M2,M3) for each landmark was calculated and the mean (Ma) of the three quantified. The percentage average deviation from Ma of the assessors was calculated as:

  • image

The percentage reliability = 100 − d. Hence the overall percentage inter-assessor reliability over the seven landmarks was

  • image

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Author contributions
  8. References

The overall intra-assessor and inter-assessor reliabilities were calculated to be 98.2 and 98.7% for the glenoid component, and 96.2 and 96.5% for the humeral component, respectively. The specific intra-assessor repeatability and inter-assessor accuracy for all the landmarks are shown in Table 1. This also shows the overall average measurements of each landmark from the reference point of the bone component. It typically took an average of 42.8 s for an assessor to select a protractor size, identify and measure a landmark. The fovea capitis was, on the overall average, located 28.5° anteriorly from the superior-central midpoint of the anatomical neck.

Table 1.   Average location, intra- and inter-assessor reliabilities for the various anatomical landmarks.
Landmarkbcdfghi
Location relative to a or e (°)28.5 ± 0.971.1 ± 6.9183.1 ± 2.633.7 ± 0.386.6 ± 0.7172.7 ± 4.7276.5 ± 7.1
Intra-assessor repeatability (%)94.495.598.797.598.298.498.7
Inter-assessor accuracy (%)97.892.798.999.499.498.198.1

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Author contributions
  8. References

The use of glenoid and humeral head protractors for the quantification of anatomical landmark is not a known method. It was therefore necessary for the present study to adopt a rigorous method of determining the reliability of its usage. The five manufactured sizes of the glenoid and HH components ensured that an appropriate fitting tool was used for each measurement and the decision as to which size to use was made independently by each assessor. Again the generalised definition of the landmarks to measure ensured the integration of the individual assessor’s judgement to complement the standardised protocol of identification of the protractor origin on the bones. Most of the landmarks to be quantified were individually broad regions of interest requiring that each assessor use their intuition to decide on the spot that represented the midsubstance of the regions, for example, the lesser tubercle and the directional margins of the glenoid. A minimum of 1 h recreation after each randomised trial was designed to establish a fresh frame of mind in order to introduce possible effects of distraction in an assessor’s judgements. We believe that this integrated possible difficulties in the use of the developed instrument.

The result shows very low intra-assessor errors when repeating the quantification of the same landmarks. The average inter-assessor accuracies of 98.7 and 96.5% for the components show that the differences in the measurements of the assessors are insignificant. The short measurement time shows that the tool is fairly easy to use. A search for literature-based instrumentation for similar angular quantifications of the GHJ hard tissues for comparison did not produce any results. However, a study that applied vector analysis technique to quantify the average angular position of the humeral fovea capitis using 21 humeri demonstrated that this is located on the anatomical neck circle-fit, 27° anteriorly from the mid-point of its most superior aspect for an erect humerus (Amadi & Bull, 2011). This vector analysis work was able to produce a high resolution of up to 0.01° as this was based on mathematical formulations. However, the present instrument can only provide a measurement resolution of 2°. Therefore the present average measurement of 28.5° for this point is comparable to the 27° quantified by Amadi & Bull (2011). Comparison was carried out for the fovea capitis only, as there was no other similar quantification of any landmark to compare from the literature.

The use of these protractors would improve the accuracy of the quantification of anatomical origins and insertions of the glenohumeral intra-capsular structures. This application will contribute an added precision in the reporting of anatomical locations of bony features of interest around the glenoid or HH of specimens in the laboratory. Further anatomical studies are therefore planned to use this for the quantification of soft tissue relationships around the GHJ, which may better inform pathology identification and surgical planning. Digital image versions of the protractor could be superimposed on medical images for quantifications, too. In the future this application might assist in the location of the absolute points of pathologies and planning surgical intervention of glenohumeral instability. In addition to pre-operative planning, the present protractor could be autoclaved and applied intra-operatively during surgical repairs to ensure a more accurate realisation of the plan. It is also hoped that some image manipulation techniques could be developed to import the numerical version into real-time arthroscopic surgical environment to assist in locating the absolute points of surgical constructions.

Variation in bone shape, osteoarthritis and pathological osseous fragmentation might possibly affect the protractor’s reliability for a repeated measure of the same spot for a landmark, especially when this has become less pronounced than normal. We did not focus particularly on the effect of pathologies in this study and non-pathological glenoid shape variations were not considered during specimen selection. Therefore any possible effects of this across the studied specimens have been integrated in the final result and can be said to be insignificant by implication. The protractor could rather be of assistance as a guide when estimating the possible location of an eroded landmark during surgical reconstructions.

Author contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Author contributions
  8. References

The authors have played significant roles at different segments of this study including concept and design, acquisition of data, data analysis/interpretation, drafting of the manuscript, manuscript reviews/readiness and approval of the present state of the article.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Author contributions
  8. References