With the development of artificial intelligence technology and the popularity of intelligent production projects, intelligent inspection systems have gradually become a hot topic in the industrial field. As a fundamental problem in the field of computer vision, how to achieve object detection in the industry while taking into account the accuracy and real-time detection is an important challenge in the development of intelligent detection systems. The detection of defects on steel surfaces is an important application of object detection in the industry. Correct and fast detection of surface defects can greatly improve productivity and product quality. To this end, this paper introduces the MSFT-YOLO model, which is improved based on the one-stage detector. The MSFT-YOLO model is proposed for the industrial scenario in which the image background interference is great, the defect category is easily confused, the defect scale changes a great deal, and the detection results of small defects are poor. By adding the TRANS module, which is designed based on Transformer, to the backbone and detection headers, the features can be combined with global information. The fusion of features at different scales by combining multi-scale feature fusion structures enhances the dynamic adjustment of the detector to objects at different scales. To further improve the performance of MSFT-YOLO, we also introduce plenty of effective strategies, such as data augmentation and multi-step training methods. The test results on the NEU-DET dataset show that MSPF-YOLO can achieve real-time detection, and the average detection accuracy of MSFT-YOLO is 75.2, improving about 7% compared to the baseline model (YOLOv5) and 18% compared to Faster R-CNN, which is advantageous and inspiring.
Background: The “killer turn” effect after posterior cruciate ligament (PCL) reconstruction is a problem that can lead to graft laxity or failure. Solutions for this situation are currently lacking. Purpose: To evaluate the clinical outcomes of a modified procedure for PCL reconstruction and quantify the killer turn using 3-dimensional (3D) computed tomography (CT). Study design: Case series; Level of evidence, 4. Methods: A total of 15 patients underwent modified PCL reconstruction with the tibial aperture below the center of the PCL footprint. Next, 2 virtual tibial tunnels with anatomic and proximal tibial apertures were created on 3D CT. All patients were assessed according to the Lysholm score, International Knee Documentation Committee (IKDC) Subjective Knee Evaluation Form, Tegner score, side-to-side difference (SSD) in tibial posterior translation using stress radiography, and 3D gait analysis. Results: The modified tibial tunnel showed 2 significantly gentler turns (superior, 109.87° ± 10.12°; inferior, 151.25° ± 9.07°) compared with those reconstructed with anatomic (91.33° ± 7.28°; P < .001 for both comparisons) and proximal (99° ± 7.92°; P = .023 and P < .001, respectively) tibial apertures. The distance from the footprint to the tibial aperture was 16.49 ± 3.73 mm. All patient-reported outcome scores (mean ± SD) improved from pre- to postoperatively: Lysholm score, from 46.4 ± 18.87 to 83.47 ± 10.54 ( P < .001); Tegner score, from 2.47 ± 1.85 to 6.07 ± 1.58 ( P < .001); IKDC sports activities score, from 19 ± 9.90 to 33.07 ± 5.35 ( P < .001); and IKDC knee symptoms score, from 17.87 ± 6.31 to 25.67 ± 3.66 ( P < .001). The mean SSD improved from 9.15 ± 2.27 mm preoperatively to 4.20 ± 2.31 mm postoperatively ( P < .001). The reconstructed knee showed significantly more adduction (by 1.642°), less flexion (by 1.285°), and more lateral translation (by 0.279 mm) than that of the intact knee ( P < .001 for all). Conclusion: Lowering the tibial aperture during PCL reconstruction reduced the killer turn, and the clinical outcomes remained satisfactory. However, SSD and clinical outcomes were similar to those of previously described techniques using an anatomic tibial tunnel.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.