Search

Due to rapid developments in the deep learning model, artificial intelligence (AI) models are expected to enhance clinical diagnostic ability and work efficiency by assisting physicians. Therefore, many hospitals and private companies are competing to develop AI-based automatic diagnostic systems using medical images. In the near future, many deep learning-based automatic diagnostic systems would be used clinically. However, the possibility of adversarial attacks exploiting certain vulnerabilities of the deep learning algorithm is a major obstacle to deploying deep learning-based systems in clinical practice. In this paper, we will examine in detail the kinds of principles and methods of adversarial attacks that can be made to deep learning models dealing with medical images, the problems that can arise, and the preventive measures that can be taken against them. [ABSTRACT FROM AUTHOR]

Purpose: This study examined the radiological and clinical outcomes of tibiotalocacalcaneal arthrodesis using retrograde intramedullary nailing in a severe hindfoot deformity and ankle/subtalar arthritis. Materials and Methods: A total of 22 patients (22 cases) with a severe hindfoot deformity and arthritis underwent tibiotalocalcaneal arthrodesis with retrograde intramedullary nails. The average age was 57.4 years (22–82 years) and the mean follow-up was 29.6 months (12–74 months). The radiological outcomes included an assessment of the preoperative and postoperative coronal ankle alignment, hindfoot alignment, sagittal alignment, and postoperative union time. The clinical outcomes were evaluated using the visual analogue scale (VAS), American Orthopaedic Foot and Ankle Society (AOFAS) score, and postoperative complications. The results were analyzed statistically by dividing the patients into two groups based on a 10° angle of deformity. Results: Regarding the preoperative coronal ankle alignment, 14 patients had a mean varus deformity of 17.8°±14.5° and six had a mean valgus deformity of 8.1°±6.6°. Postoperatively, a satisfactory postoperative coronal ankle alignment of less than 5° was obtained in all patients. Regarding the preoperative hindfoot alignment, 12 patients showed a mean varus deformity of 15.2°±10.5° and six had a mean valgus deformity of 8.1°±4.2°. In total, 94.4% (17 patients) had satisfactory postoperative hindfoot alignment of less than 5°. Radiological union was achieved in 90.9% at an average of 19.2 weeks (12–32 weeks) and there were 2 cases of nonunion. The clinical outcomes showed improvement in the mean VAS and AOFAS scores (p＜0.001, p＜0.001, respectively). Even a preoperative severe deformity more than 10° showed a significant deformity correction of coronal ankle alignment and hindfoot alignment, postoperatively (p＜0.001, p＜0.001, respectively). No significant differences were found between the patients with a preoperative coronal ankle deformity more than 10° and those less than 10° regarding the mean postoperative coronal ankle alignment (p=0.162). Conclusion: Tibiotalocalcaneal arthrodesis using retrograde intramedullary nailing is an acceptable technique for achieving satisfactory deformity correction, high union rate with minimal complications, and improvement of the clinical outcomes. In addition, tibiotalocalcaneal arthrodesis using retrograde intramedullary nailing is considered an effective treatment option, particularly in severe ankle and hindfoot deformities. [ABSTRACT FROM AUTHOR]

Purpose: To compare the results of two different instruments made of stainless steel and titanium alloy for correction of single thoracic adolescent idiopathic scoliosis (AIS) using pedicle screw instrumentation. Materials and Methods: A total of 141 patients with single thoracic AIS treated with pedicle screw instrumentation and selective thoracic fusion were retrospectively reviewed after a follow-up of 2 years. The patients had a main thoracic curve of 40° to 75° and were divided into two groups based on instrument materials; S group (stainless steel, n=90) and T group (titanium alloy, n=51). The diameter of the stainless steel rod used was 7.0 mm while that of the titanium alloy rod was 6.35 mm or 6.0 mm. Standing long-cassette radiographic measurements including various coronal and sagittal parameters for the preoperative, early postoperative and 2-year postoperative followup were analyzed. There were no significant differences in the preoperative curve characteristics between the two groups. Results: In the S group, the preoperative main thoracic curve of 51.3°±8.4° was improved to 19.0°±7.6° (63.1% correction) and the lumbar curve of 32.3°±8.4° spontaneously decreased to 12.7°±8.2° (62.9% correction) at 2 years postoperatively. In the T group, the preoperative main thoracic curve of 49.5°±8.4° and the lumbar curve of 30.3°±8.9° was improved to 18.8°±7.4° (62.2% correction) and 11.3°±5.4° (63.3% correction), respectively. The corrections of coronal curves were not statistically different between the two groups (p>0.05). The thoracic kyphosis was changed from 16.8°±8.5° to 24.3°±6.1° in the S group and from 19.6°±11.2° to 26.6°±8.5° in the T group. There were no significant differences in the changes of sagittal curves, coronal and sagittal balances at the 2-year follow-up and the number of fused segments and used screws between the two groups (p>0.05). Conclusion: When conducting surgery for single thoracic AIS using pedicles screw instrumentation, two different instruments made of stainless steel and titanium alloy showed similar corrections for coronal and sagittal curves. [ABSTRACT FROM AUTHOR]