This dissertation addresses the problem of estimating the 6D pose of knee arthroscopic surgical instruments using only RGB video. Arthroscopy is a minimally invasive procedure where the instruments appear with strong symmetries, specular reflections, and frequent occlusions, making pose estimation highly challenging. The work integrates a markerless pipeline that combines a object detector, a dense correspondence method, and a geometric pose solver, with distortion-aware processing throughout the pipeline to handle the radial distortion of arthroscopic cameras. The pipeline was trained and tested on datasets of three instruments, with evaluation based on standard pose metrics and clinically relevant measures such as tip position and axis orientation errors. The results show reliable detection performance in most test sets, and competitive pose accuracy across instruments. Mean tip error remains within millimetric precision, and mean axis orientation error is generally below 5°. Despite these promising results, runtime analysis indicates that the current throughput of about 7.5 FPS is below the requirements for real-time surgical navigation. Overall, the study demonstrates that deep learning methods are paving the way to provide accurate and robust pose estimation, competing with physical markers approaches under occlusion and symmetry conditions, supporting their potential for integration into next-generation navigation systems in knee arthroscopy. Future improvements in dataset quality, processing speed, and method selection are expected to further enhance the practical deployment of this approach.

Elbow arthroscopy was performed in 30 dogs of different breeds. The procedure was performed bilaterally in 20 of these dogs, yielding a total of 50 joints. Different lesions were found, varying from cartilage fissures (8) to fragmentation (42) of medial coronoid process (FCP) of the ulna. Osteochondritis dissecans (OCD) of the humerus medial condyle was associated in four of them. All of these cases displayed varying degrees of synovitis. Osteoarthrosis (OA) in varying intensity was observed in 44 joints. The majority of cases were treated two to four months after the manifestation of clinical signs. Good clinical recovery occurred in dogs with minimal joint lesions, where these were diagnosed and treated within four weeks of the onset of clinical symptoms. Early diagnosis and arthroscopic treatment prevent osteoarthrosis and preserve locomotor function.

ABSTRACT This study aimed to characterize the clinical and arthroscopic presentations of horses with osteochondral fragmentation at the metacarpophalangeal and metatarsophalangeal joints at the Veterinary Hospital of the University of São Paulo in a 10-year period (2010-2019). The hind limbs were affected similarly to the plantar and dorsal fragments from the first phalanx, while the thoracic limbs had a more frequent occurrence on the dorsal aspect of the joint. The Brazilian Sport Horses were the most frequently observed breed. The degree of lameness was variable and might be related to fragment localization, animal age, and physical activity. There was a significant statistical difference between the joint effusion and degree of lameness; therefore, it can be considered that the joint effusion is related to degenerative joint processes and might be triggered by the presence of the osteochondral fragment. Surgical correction, especially when performed in young animals using an arthroscopic procedure is probably the best treatment option for most cases, possibly acting as a minimizer of the occurrence of secondary joint degeneration. Prospective studies of specific breeds with a larger number of animals can expand this clinical characterization and enable a better understanding of the animal’s recovery after surgical or conservative management.

Abstract Objectives To analyze the functional outcomes in patients submitted to videoarthroscopic surgical treatment for compleat rotator cuff tears of the shoulder, with a minimum follow-up of 10 years. Methods A total of 63 patients (63 shoulders) underwent videoarthroscopic surgical repair for compleat rotator cuff tears with a minimum follow-up of 10 years. The postoperative functional outcomes of these patients were evaluated using the Constant and University of California at Los Angeles (UCLA) scores. Results The functional evaluation revealed mean UCLA and Constant scores of 26 and 93 points, respectively. Ninety-one percent of the subjects had satisfactory Constant scores, whereas 62% presented satisfactory UCLA scores. Conclusion The arthroscopic repair of rotator cuff complete tear was effective even in the long term (minimum follow-up period of 10 years). The age of the patients before surgery, size of the lesion, the degree of fatty infiltration, and evaluation of muscle trophism are important predictors of prognosis.

Abstract Objective To evaluate the functional outcomes of patients diagnosed with femoroacetabular impingement (FAI) older than 60 years, compared with those of patients of age 40 years or younger. Methods This was a retrospective review of patients with FAI who underwent hip arthroscopy between 2010 and 2015. The patients were adults aged over 60 years with Tönnis ≤ 1 matched in a 1:1 ratio with adults aged 40 years or younger, according to the type of deformity (cam, pincer, or mixed), sex, and the date when the surgery was performed. Results Thirty-four patients were included in each group. The mean age was 30.6 ± 6.9 years and 65.6 ± 4.6 years in the control and case groups, respectively. There were no significant differences between the groups at 1-year follow-up (p > 0.05). In the group with older patients (case group), we observed a change in the total Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score from 46.3 to 22.0 in the 1st postoperative year, while the control cases improved in the WOMAC score from 38.1 to 7.2 in relation to the preoperative stage. Conclusion In the group of patients ≤ 40 years old, a considerable change was observed in the WOMAC score without a statistical significance compared with the > 60 years group. This observation suggests that hip arthroscopy is beneficial when there is an appropriate selection of patients with FAI, regardless the age of the patient.

Abstract Objective To evaluate the clinical and radiographic results as well as complications related to patients undergoing arthroscopic treatment of subspine hip impingement. Methods We retrospectively evaluated 25 patients (28 hips) who underwent arthroscopic treatment of subspine impingement between January 2012 and June 2018. The mean follow-up was 29.5 months, and the patients were evaluated clinically by using the Harris hip score modified by Byrd (MHHS), the non-arthritic hip score (NAHS), and in terms of internal rotation and hip flexion. In addition, the following items were evaluated by imaging exams: the center-edge (CE) acetabular angle, the Alpha angle, the presence of a sign of the posterior wall, the degree of arthrosis, the presence of heterotopic hip ossification, and the Hetsroni classification for subspine impingement. Results There was an average postoperative increase of 26.9 points for the MHHS, 25.4 for the NAHS (p < 0.0001), 10.5° in internal rotation (p < 0.0024), and 7.9° for hip flexion (p < 0.0001). As for the radiographic evaluation, an average reduction of 3.3° in the CE angle and of 31.6° for the Alpha angle (p < 0.0001). Eighteen cases (64.3%) were classified as grade 0 osteoarthritis of Tönnis, and 10 (35.7%) were classified as Tönnis grade 1. Two cases (7.1%) presented grade 1 ossification of Brooker. Most hips (n = 15, 53.6%) were classified as type II of Hetsroni et al. Conclusion In the present study, patients undergoing arthroscopic treatment with subspine impingement showed improvement in clinical aspects and radiographic patterns measured postoperatively, with an average follow-up of 29.5 months.

Abstract Objective To evaluate the influence of the supraspinal tear pattern on the pre- and postoperative functional evaluations. Methods A retrospective cohort study comparing patients with supraspinatus crescent-shaped tears versus L- or U-shaped tears. We included patients undergoing complete supraspinatus arthroscopic repair. We did not include patients with subscapularis or infraspinatus repair, those submitted to open surgery, or those in whom only partial repair was achieved. The clinical scales used were the American Shoulder and Elbow Surgeons Standardized Shoulder Assessment (ASES) and the Modified-University of California at Los Angeles Shoulder Rating Scale (UCLA), which were applied 1 week before and 24 months after the procedure. Results We analyzed 167 shoulders (from 163 patients). In the preoperative period, the ASES scale was significantly higher in the crescent-shaped pattern (43.5 ± 17.6 versus 37.7 ± 13.8; p = 0.034). The UCLA scale followed the same pattern (15.2 ± 4.6 versus 13.5 ± 3.6; p = 0.028). In the postoperative period, however, there was no significant difference. According to the ASES scale, crescent-shaped tears scored 83.7 ± 18.7 points, and L- or U-shaped tears scored 82.9 ± 20.1 (p = 0.887). The values were 30.9 ± 4.9 and 30.5 ± 5.6 (p = 0.773) respectively, by the UCLA scale. Conclusion Crescent-shaped and L- or U-shaped supraspinatus tears have similar postoperative functional results. In the preoperative period, the functional results are superior in crescent-shaped tears.

Abstract Slipped capital femoral epiphysiolysis (SCFE) may result in femoroacetabular impingement (FAI) of the hip in up to one third of the cases. Residual deformity of the cam-type, or “pistol-grip”, is associated with chondrolabral injury, resulting in pain, functional disability, and early osteoarthritis. The arthroscopic treatment with osteochondroplasty proved to be beneficial in a selected case of FAI secondary to SCFE.

Abstract Objective The open Bristow procedure is a long established and effective method for treating anterior shoulder instability. Following the trends of minimally-invasive surgeries, these procedures were performed arthroscopically, and their outcomes were evaluated. Methods A total of 43 shoulders of patients submitted to Bristow procedures by arthroscopy, using a graft positioned horizontally and a screw, with at least two years of postoperative follow-up, were evaluated regarding quality of life, de novo dislocation index, and loss of lateral rotation. Results The mean follow-up time was of 76 months (range: 129 to 24 months). The University of California at Los Angeles (UCLA) score varied from 25.56 ± 0.50 (standard deviation [SD] = 3.25) to 33.23 ± 0.44 (SD = 2.91) (p < 0.0001). Two or more years after surgery, the mean Rowe score was of 94.25 ± 1.52 (SD = 1.34), whereas the good results standard is 75 (p < 0.0001). The mean value for the simple shoulder test was of 11.35 ± 0.21 (SD = 1.34), while the mean value of the lateral rotation loss was of 10.37º ± 1.36º (SD = 8.58º). There were no de novo dislocations. In total, there were 12 complications, 8 of which had no clinical repercussions. The clinically-significant complications included an infection six months after surgery with a potential hematogenous origin, a coracoid fracture that required an intraoperatively procedure change, and two patients with previous impingement who required synthesis material removal more than six months after surgery. Conclusion Although the arthroscopic Bristow procedure was effective in treating anterior shoulder instability, it is not a complication-free surgery.

Abstract Objectives To evaluate the clinical results of patients submitted to arthroscopic treatment of partial lesion of the articular part of the rotator cuff by transtendon suture techniques and after completing the lesion and to compare the postoperative recovery time of the two techniques. Method Retrospective study based on the identification of all cases with partial lesion of the articular part of the rotator cuff submitted to arthroscopic treatment from October 1999 to December 2016 at the Shoulder and Elbow Group of our institution. Thirty-nine patients were included and divided intotwo groups: those who underwent the transtendon technique and those in whom the lesion was completed. The two groups were statistically similar. The respective medical records were analyzed as well as the identification of the applied technique, the postoperative results, and the related complications. The functional evaluation was performed using the score of the University of California at Los Angeles (UCLA). Results There was no statistically relevant difference between the groups, with a mean UCLA score of 32, and no difference in the time required for rehabilitation. Conclusion There was no difference between the clinical outcome of the patients, regardless of the technique used to repair the partial lesions of the articular part of the rotator cuff, with satisfactory results in 93% of the cases.