Background and purpose — We aimed to evaluate the ability of artificial intelligence (a deep learning algorithm) to detect and classify proximal humerus fractures using plain anteroposterior shoulder radiographs.Patients and methods — 1,891 images (1 image per person) of normal shoulders (n = 515) and 4 proximal humerus fracture types (greater tuberosity, 346; surgical neck, 514; 3-part, 269; 4-part, 247) classified by 3 specialists were evaluated. We trained a deep convolutional neural network (CNN) after augmentation of a training dataset. The ability of the CNN, as measured by top-1 accuracy, area under receiver operating characteristics curve (AUC), sensitivity/specificity, and Youden index, in comparison with humans (28 general physicians, 11 general orthopedists, and 19 orthopedists specialized in the shoulder) to detect and classify proximal humerus fractures was evaluated.Results — The CNN showed a high performance of 96% top-1 accuracy, 1.00 AUC, 0.99/0.97 sensitivity/specificity, and 0.97 Youden index for distinguishing normal shoulders from proximal humerus fractures. In addition, the CNN showed promising results with 65–86% top-1 accuracy, 0.90–0.98 AUC, 0.88/0.83–0.97/0.94 sensitivity/specificity, and 0.71–0.90 Youden index for classifying fracture type. When compared with the human groups, the CNN showed superior performance to that of general physicians and orthopedists, similar performance to orthopedists specialized in the shoulder, and the superior performance of the CNN was more marked in complex 3- and 4-part fractures.Interpretation — The use of artificial intelligence can accurately detect and classify proximal humerus fractures on plain shoulder AP radiographs. Further studies are necessary to determine the feasibility of applying artificial intelligence in the clinic and whether its use could improve care and outcomes compared with current orthopedic assessments.
Magnetic resonance (MR) imaging is a powerful diagnostic tool that can be used to help evaluate spinal infection and to help distinguish between an infection and other clinical conditions. In most cases of spinal infection, MR images show typical findings such as vertebral endplate destruction, bone marrow and disk signal abnormalities, and paravertebral or epidural abscesses. However, it is not always easy to diagnose a spinal infection, particularly when some of the classic MR imaging features are absent or when there are unusual patterns of infectious spondylitis. Furthermore, noninfectious inflammatory diseases and degenerative disease may simulate spinal infection. It is necessary to be familiar with atypical MR imaging findings of spinal infection and features that may mimic spinal infection to avoid misdiagnosis and inappropriate treatment.
BackgroundMRI analysis of subtalar ligaments in the tarsal sinus has not been well performed. We retrospectively investigated the appearance of subtalar ligaments using 3D isotropic MRI and compared imaging findings of subtalar ligaments between patients with subtalar instability (STI) and controls.MethodsPreoperative MRIs of 23 STI patients treated with arthroscopic subtalar reconstruction were compared to MRIs of 23 age- and sex-matched control subjects without STI. Thickness and width of anterior capsular ligament (ACL) and interosseous talocalcaneal ligament (ITCL) as well as thickness of calcaneofibular ligament (CFL) and anterior talofibular ligament (ATFL) were measured. Abnormalities in ACL, ITCL, CFL, ATFL, cervical ligament, and inferior extensor retinaculum were analyzed.ResultsSTI patients had significantly smaller ACL thickness and ACL width than controls (ACL thickness: 1.73 mm vs. 2.22 mm, p = 0.007; ACL width: 7.21 mm vs. 8.80 mm, p = 0.004). ACL thickness of ≤2.1 mm had a sensitivity of 66.7% and a specificity of 66.7% for diagnosis of STI. ACL width of ≤7.9 mm had a sensitivity of 80.0% and a specificity of 76.2% for the diagnosis of STI. However, thickness and width of ITCL, thickness of CFL, or thickness of ATFL was not significantly different between the two groups. Absence or complete tear of ACL was significantly more frequent in STI patients than that in controls (34.8% vs. 8.7%, p = 0.035). Complete tear of CFL and ATFL was more common in STI patients than that in controls, although the difference between the two groups was not statistically significant. Abnormalities of ITCL, cervical ligament, or inferior extensor retinaculum were not significantly different between the two groups.ConclusionsMRI features of thin or narrow ACLs may suggest STI. Absence or complete tear of ACL was significantly more common in STI patients than that in controls.
ObjectiveTo evaluate the usefulness of time-resolved contrast enhanced magnetic resonance angiography (4D MRA) after stent-assisted coil embolization by comparing it with time of flight (TOF)-MRA.Materials and MethodsTOF-MRA and 4D MRA were obtained by 3T MRI in 26 patients treated with stent-assisted coil embolization (Enterprise:Neuroform = 7:19). The qualities of the MRA were rated on a graded scale of 0 to 4. We classified completeness of endovascular treatment into three categories. The degree of quality of visualization of the stented artery was compared between TOF and 4D MRA by the Wilcoxon signed rank test. We used the Mann-Whitney U test for comparing the quality of the visualization of the stented artery according to the stent type in each MRA method.ResultsThe quality in terms of the visualization of the stented arteries in 4D MRA was significantly superior to that in 3D TOF-MRA, regardless of type of the stent (p < 0.001). The quality of the arteries which were stented with Neuroform was superior to that of the arteries stented with Enterprise in 3D TOF (p < 0.001) and 4D MRA (p = 0.008), respectively.Conclusion4D MRA provides a higher quality view of the stented parent arteries when compared with TOF.
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.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.