Three-dimensional geometry of the normal shoulder: a software analysis

BACKGROUND

Three-dimensional (3D) geometry of the normal glenohumeral bone anatomy and relations is poorly documented. Our aims were (1) to determine the 3D geometry of the normal glenohumeral joint (GHJ) with reference to the scapular body plane and (2) to identify spatial correlations between the orientation and direction of the humeral head and the glenoid.

METHODS

Computed tomographies (CTs) of the normal, noninjured GHJ were collected from patients who had undergone CTs in the setting of (1) polytrauma, (2) traumatic head injury, (3) chronic acromioclavicular joint dislocations, and (4) unilateral trauma with a contralateral normal shoulder. We performed 3D segmentation and measurements with a fully automatic software (Glenosys; Imascap). Measurements were made in reference to the scapular body plane and its transverse axis. Geometric measurements included version, inclination, direction, orientation, best-fit sphere radius (BFSR), humeral subluxation, critical shoulder angle, reverse shoulder angle, glenoid area, and glenohumeral distance. Statistical correlations were sought between glenoid and humeral 3D measurements (Pearson correlation).

RESULTS

A total of 122 normal GHJs (64 men, 58 women, age: 52 ± 17 years) were studied. The glenoid BFSR was always larger than the humerus BFSR (constant factor of 1.5, standard deviation = 0.2). The mean glenoid version and inclination were -6° ± 4° and 7° ± 5°, respectively. Men and women were found to have significantly different values for inclination (6° vs. 9°, P = .02), but not for version. Humeral subluxation was 59% ± 7%, with a linear correlation with glenoid retroversion (r = -0.70, P < .001) regardless of age. There was a significant and linear correlation between glenoid and humeral orientation and direction (r = 0.72 and r = 0.70, P < .001).

CONCLUSION

The 3D geometry of the glenoid and humeral head present distinct limits in normal shoulders that can be set as references in daily practice: version and inclination are -6° and 7°, respectively, and humeral posterior subluxation is 59%; interindividual variations, regardless of the size, are relative to the scapular plane. There exists a strong correlation between the position of the humeral head and the glenoid orientation and direction.

Evaluation of three-dimensional acromiohumeral distance in the standing position and comparison with its conventional measuring methods

Background

Narrowing of the acromiohumeral distance (AHD) implies a rotator cuff tear. However, conventional AHD measurements using two-dimensional (2D) imaging or with the patient in the supine position might differ from that while standing during daily activity. This study aimed to evaluate the three-dimensional (3D) actual distance between the acromion and humeral head in the standing position and compare the AHD values with those obtained using conventional measuring methods.

Methods

Computed tomography (CT) images of 166 shoulders from 83 healthy volunteers (31 male and 52 female; mean age 40.1 ± 5.8 years; age range, 30–49 years) were prospectively acquired in the supine and standing positions using conventional and upright CT scanners, respectively. The minimum distance between the acromion and humeral head on the 3D surface models was considered as the 3D AHD. We measured the 2D AHD on anteroposterior digitally reconstructed radiographs. The AHD values were compared between the supine and standing positions and between the 2D and 3D measurements.

Results

The mean values of 2D AHD were 8.8 ± 1.3 mm (range, 5.9–15.4 mm) in the standing position and 8.1 ± 1.2 mm (range, 5.3–14.3 mm) in the supine position. The mean values of 3D AHD were 7.3 ± 1.4 mm (range, 4.7–14.0 mm) in the standing position and 6.6 ± 1.2 mm (range, 4.4–13.7 mm) in the supine position. The values of 3D AHD were significantly lower than those of 2D AHDs in both the standing and supine positions (P < 0.001). The values of 2D and 3D AHDs were significantly lower in the supine position than in the standing position (P < 0.001).

Conclusions

This study evaluated the 3D AHD of normal shoulders in the standing position using an upright CT scanner. The present results indicated that assessments in the supine position can underestimate the value of the AHD compared with those made in the standing position and that assessments using 2D analysis can overestimate the value.