The purpose of this study was to develop a convolutional neural network (CNN) for fracture detection, classification, and identification of greater tuberosity displacement ≥ 1 cm, neck-shaft angle (NSA) ≤ 100°, shaft translation, and articular fracture involvement, on plain radiographs. The CNN was trained and tested on radiographs sourced from 11 hospitals in Australia and externally validated on radiographs from the Netherlands. Each radiograph was paired with corresponding CT scans to serve as the reference standard based on dual independent evaluation by trained researchers and attending orthopaedic surgeons. Presence of a fracture, classification (non- to minimally displaced; two-part, multipart, and glenohumeral dislocation), and four characteristics were determined on 2D and 3D CT scans and subsequently allocated to each series of radiographs. Fracture characteristics included greater tuberosity displacement ≥ 1 cm, NSA ≤ 100°, shaft translation (0% to < 75%, 75% to 95%, > 95%), and the extent of articular involvement (0% to < 15%, 15% to 35%, or > 35%).Aims
Methods
Accurate characterisation of fractures is essential in fracture management trials. However, this is often hampered by poor inter-observer agreement. This article describes the practicalities of defining the fracture population, based on the Neer classification, within a pragmatic multicentre randomised controlled trial in which surgical treatment was compared with non-surgical treatment in adults with displaced fractures of the proximal humerus involving the surgical neck. The trial manual illustrated the Neer classification of proximal humeral fractures. However, in addition to surgical neck displacement, surgeons assessing patient eligibility reported on whether either or both of the tuberosities were involved. Anonymised electronic versions of baseline radiographs were sought for all 250 trial participants. A protocol, data collection tool and training presentation were developed and tested in a pilot study. These were then used in a formal assessment and classification of the trial fractures by two independent senior orthopaedic shoulder trauma surgeons.Objectives
Methods
The purpose of this study was to review the long-term outcomes of a previously reported prospective series of 46 type III acromioclavicular dislocations. These were treated surgically with temporary fixation of the acromioclavicular joint with wires, repair of the acromioclavicular ligaments, and overlapped suture of the deltoid and trapezius muscles. Of the 46 patients, one had died, four could not be traced, and three declined to return for follow-up, leaving 38 patients in the study. There were 36 men and two women, with a mean age at follow-up of 57.3 years (41 to 71). The mean follow-up was 24.2 years (21 to 26). Patients were evaluated using the Imatani and University of California, Los Angeles (UCLA) scoring systems. Their subjective status was assessed using the Disabilities of the Arm, Shoulder and Hand and Simple Shoulder Test questionnaires, and a visual analogue scale for patient satisfaction. The examination included radiographs of the shoulder. At a follow-up of 21 years, the results were satisfactory in 35 (92.1%) patients and unsatisfactory in three (7.9%). In total, 35 patients (92.1%) reported no pain, one slight pain, and two moderate pain. All except two patients had a full range of shoulder movement compared with the opposite side. Unsatisfactory results were the result of early redisplacement in two patients, and osteoarthritis without redisplacement in one. According to the Imatani and UCLA scores, there was no difference between the operated shoulder and the opposite shoulder (p >
0.05). Given the same situation, 35 (92.1%) patients would opt for the same surgical treatment again. Operative treatment of type III acromioclavicular joint injuries produces satisfactory long-term results.
We treated 47 patients with a mean age of 57 years (22 to 88) who had a proximal humeral fracture in which there was a severe varus deformity, using a standard operative protocol of anatomical reduction, fixation with a locking plate and supplementation by structural allografts in unstable fractures. The functional and radiological outcomes were reviewed. At two years after operation the median Constant score was 86 points and the median Disabilities of the Arm, Shoulder and Hand score 17 points. Seven of the patients underwent further surgery, two for failure of fixation, three for dysfunction of the rotator cuff, and two for shoulder stiffness. The two cases of failure of fixation were attributable to violation of the operative protocol. In the 46 patients who retained their humeral head, all the fractures healed within the first year, with no sign of collapse or narrowing of the joint space. Longer follow-up will be required to confirm whether these initially satisfactory results are maintained.
This randomised study compared outcomes in patients with displaced fractures of the clavicle treated by open reduction and fixation by a reconstruction plate which was placed either superiorly or three-dimensionally. Between 2003 and 2006, 133 consecutive patients with a mean age of 44.2 years (18 to 60) with displaced midshaft fractures of the clavicle were allocated randomly to a three-dimensional (3D) (67 patients) or superior group (66). Outcome measures included the peri-operative outcome index, delayed union, revision surgery and symptoms beyond 16 weeks. CT was used to reconstruct an image of each affected clavicle and Photoshop 7.0 software employed to calculate the percentage of the clavicular cortical area in the sagittal plane. The patients were reviewed clinically and radiographically at four and 12 months after the operation. The superior plate group had a higher rate of delayed union and had more symptomatic patients than the 3D group (p <
0.05). The percentage comparisons of cortical bone area showed that cortical bone in the superior distal segment is thicker than in the inferior segment, it is also thicker in the anterior mid-section than in the posterior (p <
0.05). If fixation of midshaft fractures of the clavicle with a plate is indicated, a 3D reconstruction plate is better than one placed superiorly, because it is consistent with the stress distribution and shape of the clavicle.
The generally-accepted treatment for large, displaced fractures of the glenoid associated with traumatic anterior dislocation of the shoulder is operative repair. In this study, 14 consecutive patients with large (>
5 mm), displaced (>
2 mm) anteroinferior glenoid rim fractures were treated non-operatively if post-reduction radiographs showed a centred
We used an inverted shoulder arthroplasty in 43 consecutive patients with a mean age of 78 years (65 to 97) who had sustained a three- or four-part fracture of the upper humerus. All except two were reviewed with a mean follow-up of 22 months (6 to 58). The clinical outcome was satisfactory with a mean active anterior elevation of 97° (35° to 160°) and a mean active external rotation in abduction of 30° (0° to 80°). The mean Constant and the mean modified Constant scores were respectively 44 (16 to 69) and 66% (25% to 97%). Complications included three patients with reflex sympathetic dystrophy, five with neurological complications, most of which resolved, and one with an anterior dislocation. Radiography showed peri-prosthetic calcification in 36 patients (90%), displacement of the tuberosities in 19 (53%) and a scapular notch in ten (25%). Compared with conventional hemiarthroplasty, satisfactory mobility was obtained despite frequent migration of the tuberosities. However, long-term results are required before reverse shoulder arthroplasty can be recommended as a routine procedure in complex fractures of the upper humerus in the elderly.
