Acromial impingement syndrome is one of the most common causes of shoulder pain, and it can result in rotator cuff injury. Whether the underlying cause of this shoulder pain is the degenerative change of rotator cuff tendon or the impact of anterior acromial spur on acromion is still debatable. Many studies still cannot deny that external mechanical collision is one of the main causes of rotator cuff degeneration and shoulder pain. The acromial spur has been reported to be associated with acromial impingement and resulting rotator cuff injury and is considered to be formed by the contraction of acromiocoracoid ligament. The incidence of anterior acromial spur in the "hooked" acromion is higher than those of other types of acromion. Therefore, the anterior acromial spur may have a certain relationship with the morphological changes of acromion, acromial impingement and rotator cuff injury. The preoperative and postoperative evaluations of patient's anterior acromial spur are beneficial to the diagnosis, treatment and rehabilitation guidance of shoulder joint disease.In this study, a three-dimensional morphological investigation was conducted on the preoperative and postoperative anterior acromial spurs and acromial morphology of patients who receive shoulder arthroscopic surgery. The following questions were discussed: (1) Whether the different types of acromion and the degree of rotator cuff injury were associated with acromial morphology, and whether the size of anterior acromial spur is related to age. (2) Whether the morphological changes of anterior acromial spur after acromioplasty can change the morphology of acromion, and whether it is related to the recovery of postoperative shoulder function.

1. Research object: From October 2016 to May 2018, a total of 129 cases of shoulder arthroscopy were performed in the Department of Orthopaedic Sports Medicine of the Ninth People's Hospital affiliated to Shanghai Jiaotong University. All patients underwent arthroscopic acromioplasty. The standard radiographs of anterior and posterior view, supraspinatus outlet view and thin slice CT scan were conducted before and after surgery. 2. Inclusive criteria: (1) preoperative or intraoperative diagnosis of acromial impingement; (2) shoulder joint adhesion and subacromial bursitis caused by various reasons; (3) rotator cuff injuries. 3. Exclusive criteria: (1) previous history of shoulder surgery , fracture, infection and tumor; (2) acromioclavicular arthritis or progressive glenohumeral arthritis; (3) intraoperative diagnosis of labial lesion. 4. Acromion morphological classification, surgical procedure and rotator cuff injury criteria:In 1986, Bigliani divided acromial morphologyinto three types. Type I was flat; type II was arc-like; type III was hooked. All patients undergoing surgery were divided into type I, type II and type III based on standard supraspinatus outlet radiograph. All the operations were completed by a senior associate chief physician. The patient was routinely in lateral position, and the glenohumeral joint was observed through standard posterior and anterior working approaches. The subacromial gap was explored through posterior approach.With the establishment of standard lateral approach, the size and position of rotator cuff tear were observed after the debridement of subacromial bursa. All patients underwent routine acromioplasty, and the single row fixation method was adopted if there was rotator cuff injury. The degree of rotator cuff tear was determined by preoperative MRI and arthroscopic examination, and they were divided into three groups: complete rotator cuff injury, partial rotator cuff injury and no rotator cuff injury. The complete injury of rotator cuff is defined as its full thickness injury. Ellman divided rotator cuff tear into three categories, namely, partial tear on bursal surface, interstitial tear, and partial tear on articular surface. Each category was divided into three degrees based on the depth of tear: 1st degree referred to the tear depth < 3 mm; 2nd degree referred to the tear depth between 3 to 6 mm; 3rd degree referred to the tear depth > 6 mm or more than 50% thickness of the tendon. Any injury of the above types observed under arthroscopy was defined as partial rotator cuff tear. Both the full-thickness rotator cuff injury and the 3rd degree partial rotator cuff injury were routinely performed procedures for repair or suture fixation. 5. CT scan image processing and three-dimensional modeling: The preoperative and postoperative shoulder joint scans were performed using the latest Philips 64-row multi-slice spiral CT with the scan layer thickness of 0.625 mm. The DICOM image data was imported into Medraw three-dimensional modeling software (Image Medraw Technology Co, Ltd, Shanghai, China) , and the sequence with thin layer thickness and more layers were selected for three-dimensional reconstruction; the threshold segmentation range was set from 180 to 220 HU to obtain the three-dimensional model of bone structure. The bone structure with low HU value was repaired with hand-drawing function, and the humerus, clavicle, coracoid process, etc. were distinguished and reconstructed to obtain the three-dimensional model for measurement. 6. Establishmentof model with three-dimensional coordinate system and measurement of anterior acromial osteophyte and acromial morphology:The central point A of the three-dimensional structure in glenohumeral joint, the innermost point B of scapula and the lowermost corner point C of scapula were simulated using software algorithm. The scapula plane of anteroposterior view was formed by point A, B, and C. With point A as the origin, the x-axis was formed in inward and outward direction; the y-axis was formed in superoinferior direction; the z-axis was formed in anteroposterior direction.Both the acromial spur distance (ASD) and the acromial morphology were measured on standardized lateral 3D model. The acromial slope angle (ASA) and acromial curvature (ACH) were selected as acromial morphological indicators. The ASA was defined as the angle formed by the midpoint of acromial arc to both the anterior and posterior edges on scapular lateral view. The height of acromial arc was defined as the maximum distance from the midpoint of acromial arc to the line of both anterior and posterior edges. ASD is the distance of real acromial anterior edge forward. The real acromial anterior edge and its spurs were marked on CT scan, and the distances between the corresponding points were measured on 3D model. The distances between the anterior margin of acromion and its shape parameters were obtained by the same method above.7. Postoperative shoulder function scores: All the patients who received shoulder arthroscopic rotator cuff repair were followed up in the outpatient clinic 6 months after the operations, and the shoulder function scores included Constant Shoulder Scoring System and UCLA Shoulder Scoring System.8. Statistical analysis: Data processing was performed using SPSS 13.0 software. The Kappa coefficient was used to evaluate the reliability of the measurement. One-way ANOVA and Pearson correlation coefficient (PCC) were used to determine the relationship between age, ASA, ACH, and ASD as well as the different groups of acromial morphology and degrees of rotator cuff injury. The Mann-Whitney U test was used to compare the preoperative and postoperative parameters. A P value <0.05 was considered statistically different.

CT scan data reading, shoulder joint 3D model reconstruction and data measurement were performed independently by two junior orthopaedic surgeons. The preoperative acromial morphological classification and the degree of rotator cuff injury during operation were determined by a senior surgeon. The data surveying physician was completely unaware of the sample details, and the final result was the average of the measurement results of the two physicians. The results showed that the measurement had good consistency and high credibility.There was a significant positive correlation between ASD and age with P=0.014. There was no significant correlation between ASA and age or ACH and age with P>0.05. The difference of age between the groups of various acromial classification was statistically significant. There was no significant difference between ASA and ACH in different groups of acromial classification or degrees of rotator cuff injury (P>0.05) , and the ASD showed certain differences.The shoulder arthroscopic acromioplasty significantly reduced ASD with P<0.001. However, the reduction in ASD did not lead to significant differences in ASA and ACH between the type I and type II acromion groups before and after surgery.There was significant difference in the type III acromion group, and the P values were 0.012 and 0.038, respectively. The average Constant score was (92.21±4.11) score 3 months after operation, and the average UCLA score was (30.96±2.54) score. These were not significantly correlated with the change of ASD before and after surgery (P=0.427) .

The evaluation with digitization and 3D modeling can improve the accuracy and extent of acromial morphology measurement. Age is still an essential indicator for the prediction of acromial morphology and anterior acromial osteophyte. The incidence of ASD increase is higher in the type III "hooked" acromion and can change its original morphology, causing the increases of ASA and ACH, which ultimately results in the risks of acromial impingement and rotator cuff injury. Conventional acromioplasty should be performed under arthroscopy for the patients with type III acromion, and special attention should be paid to the positions of anterior acromial spur to ensure the normal morphological restoration of acromion. For patients with type I or II acromion, arthroscopic acromioplasty is optional.