1. Inclusion and exclusion criteria
The inclusion criteria were: (1) no prior surgery and injury to the dominant shoulder; (2) no obvious surgical contra-indication in preoperative examination; (3) no contra-indication of radiological examination; (4) patient agreed to participate in the study and filled out an informed consent form.
The exclusion criteria were: (1) the primary disease was an inflammatory disease, such as giant rotator cuff tear, etc., that the researchers decided as unsuitable for inclusion; (2) patient had participated in other clinical trials within the previous three months; (3) other conditions deemed unsuitable for inclusion in this study as judged by the investigator.
2 Reconstruction of a 3D model of the proximal humerus
2.1 Acquisition of imaging data
The shoulder joint of the patient was CT scanned, with the affected shoulder placed on the side of the body, and a 3D model of the biceps groove was constructed based on CT scan data. A 64-slice CT scanner (SOMATOM Perspective, China) was used, covering the upper end of the humerus. The layer thickness was 0.625 mm, the layer spacing was 0.95 mm, and each pixel of the obtained image was 512 × 512. After scanning, the CT images were preprocessed in CT workstation, and CT data obtained were stored in the DICOM format.
2.2 Three-dimensional reconstruction of proximal humeral model
DICOM scan data were imported into Mimics 21.0 software. Threshold analysis was performed using the threshold tool, which was set at 245. The humeral boundary was isolated, the joining images were excised using the zone growth tool, and the excess structures were isolated. The layers were edited one by one, with the remnants replenished and the noise removed. The model was further optimized in Geomagic Warp reverse engineering software. The software was used to fill holes, remove noise, and repair boundaries to smooth the model, resulting in a 3D model of the humeral head containing the tackle structure, as displayed in Fig. 1.
3 Research design
3.1 Positioning the measurement plane
The depth of the internodular sulcus was defined as the distance from the highest point of the lesser nodule to the bottom of the biceps groove [11, 12]. It is critical to locate the measurement plane passing through the highest point of the tuberosity and the bottom of the biceps groove. Previous studies based on CT or MRI data were prone to errors because they only relied on imaging cross-sectional images to select the measurement plane, which could be affected by body position and the angle of administration. In this study, 3-MATIC software was used to set up a humerus model parallel to the humerus shaft Reference line (L1). The highest point of the tuberosity was selected as the reference plane S1 perpendicular to L1, and the section S2 of the humerus shaft passing through the highest point of the tuberosity was obtained. This plane was defined as the measurement plane passing through the highest point of the lesser tubercle and the sulcus floor, and measurements of the bony parameters of the internodular sulcus opening were performed on S2 as shown in Fig. 2.
3.2 Measurement of the bony structure of biceps groove
(1) Width of groove (WG)
On S2 plane, the width of the groove refers to the straight line distance between the vertices of the large and small nodules.
(2) Depth of groove (DG)
The depth of the biceps groove is the length of a straight line perpendicular to the vertex of the nodules.
(3) Opening Angle of biceps groove (OA)
The lowest point of the biceps groove was selected, and a tangent line was made along the lateral wall of the nodules. The angle between the two points was the opening angle of the biceps groove.
(4) Medial Wall Angle of biceps groove opening (MWA)
The lowest point of the biceps groove was made parallel to the vertex of the nodules, and the tangent line of the lowest point of the biceps groove along the medial wall of the small nodules was made. The angle between the two was the medial wall angle of the biceps groove, as depicted in Fig. 3.
3.3 Measurement of inclination angle of intertubercular sulcus opening
In this study, the angle between the two lines of the large and small nodules and the transverse line of the humerus shaft was defined as the opening inclination angle of the intertubercular sulcus. Since the connection of the vertices of the large and small nodules is not in the same plane in 3D, the connection L1 of vertices of the large and small nodules were directly projected, and the projected line L2 and the vertical segmentation plane S1 of humerus bone were established, in UG (Unigraphics NX Siemens USA) software. The angle between L2 and S1 was measured to obtain the opening inclination angle of the intertubercular sulcus, as demonstrated in Fig. 4.
4 Arthroscopy
Arthroscopy is considered the gold standard for evaluating pulley and LHBT injury. In this study, arthroscopy was performed by the same senior shoulder surgeon to observe the degree of LHBT injury in the glenohumeral joint and classify the pulley structure injury.
4.1 Classification of pulley structure injury
The method proposed by Martetschläger [13] was used to classify pulley structure injury, as illustrated in Fig. 5.
• Type I: medial pulley structure injury (medial coracohumeral ligament and/or superior glenohumeral ligament)
• Type II: lateral pulley structure injury (lateral coracoid brachial ligament)
• Type III: combined injury of internal and lateral pulley structure
4.2 Classification of injury of LBHT
The method proposed by Lafosse [13] was used to classify LBHT injury, as shown in Fig. 6.
• Grade 0: no injury to long head tendon of biceps
• Grade I: minor injury (less than 50% local loss or erosion of tendon)
• Class II: major injury (extensive absence or erosion of more than 50% of the tendon)
5 Statistical analysis
Statistical analyses were performed using SPSS 23.0 software. Empirical measurements with a normal distribution and homogeneity of variance were reported as mean ± standard deviation (x ̅±s), or as median and quartile spacing otherwise. Differences between groups were compared by independent sample T-test. Count data were represented with frequency table, composition ratio, etc. Correlations were performed using the Spearman correlation test. Differences less than 0.05 (P < 0.05) were considered statistically significant.