ABSTRACT
In recent years, fracture image diagnosis using a convolutional neural network (CNN) has been reported. The purpose of the present study was to evaluate the ability of CNN to diagnose distal radius fractures (DRFs) using frontal and lateral wrist radiographs. We included 503 cases of DRF diagnosed by plain radiographs and 289 cases without fracture. We implemented the CNN model using Keras and Tensorflow. Frontal and lateral views of wrist radiographs were manually cropped and trained separately. Fine-tuning was performed using EfficientNets. The diagnostic ability of CNN was evaluated using 150 images with and without fractures from anteroposterior and lateral radiographs. The CNN model diagnosed DRF based on three views: frontal view, lateral view, and both frontal and lateral view. We determined the sensitivity, specificity, and accuracy of the CNN model, plotted a receiver operating characteristic (ROC) curve, and calculated the area under the ROC curve (AUC). We further compared performances between the CNN and three hand orthopedic surgeons. EfficientNet-B2 in the frontal view and EfficientNet-B4 in the lateral view showed highest accuracy on the validation dataset, and these models were used for combined views. The accuracy, sensitivity, and specificity of the CNN based on both anteroposterior and lateral radiographs were 99.3, 98.7, and 100, respectively. The accuracy of the CNN was equal to or better than that of three orthopedic surgeons. The AUC of the CNN on the combined views was 0.993. The CNN model exhibited high accuracy in the diagnosis of distal radius fracture with a plain radiograph.
Subject(s)
Deep Learning , Orthopedic Surgeons , Humans , Neural Networks, Computer , Radiography , Wrist/diagnostic imagingABSTRACT
As fractures heal, immature callus formed in the hematoma is calcified by osteoblasts and altered to mature bone. Although the bone strength in the fracture-healing process cannot be objectively measured in clinical settings, bone strength can be predicted by specimen-specific finite element modeling (FEM) of quantitative computed tomography (qCT) scans. FEM predictions of callus strength would enable an objective treatment plan. The present study establishes an equation that converts material properties to bone density and proposes a specimen-specific FEM. In 10 male New Zealand white rabbits, a 10-mm long bone defect was created in the center of the femur and fixed by an external fixator. The callus formed in the defect was extracted after 3-6 weeks, and formed into a (5 × 5 × 5 mm3) cube. The bone density measured by qCT was related to the Young's modulus and the yield stress measured with a mechanical tester. For validation, a 10-mm long bone defect was created in the central femurs of another six New Zealand white rabbits, and fixed by an external fixator. At 3, 4, and 5 weeks, the femur was removed and subjected to Computed tomography (CT) scanning and mechanical testing. A specimen-specific finite element model was created from the CT data. Finally, the bone strength was measured and compared with the experimental value. The bone mineral density σ was significantly and nonlinearly correlated with both the Young's modulus E and the yield stress σ. The material-property conversion equations were E = 0.2391e8.00ρ and ρ = 30.49σ2.41. Moreover, the experimental bone strength was significantly linearly correlated with the prospective FEM. We demonstrated the Young's moduli and yield stresses for different bone densities, enabling a FEM of the bone-healing process. An FEM based on these material properties is expected to yield objective clinical judgment criteria.
Subject(s)
Bone Density , Bony Callus , Animals , Biomechanical Phenomena , Bony Callus/diagnostic imaging , Elastic Modulus , Finite Element Analysis , Male , Prospective Studies , Rabbits , Stress, MechanicalABSTRACT
BACKGROUND: To assess the surgical results of distal radius fractures with the involvement of a volar rim fragment using the DePuy-Synthes 2.0 mm and 2.4 mm locking plates. METHODS: Subjects were six women and one man of average age 57 years (range, 31-83 y) and a mean follow-up period of 9 months (range, 5-19 mo) with AO B3 (volar shearing) distal radius fractures. Time of the procedure, physical examination of wrist range of motion, grip strength compared with the contralateral healthy wrist, and radiographic evaluation (volar tilt, radial inclination, and ulnar variance) were evaluated. The Visual Analog Scale (VAS: 0 points represents no pain, 100 points represents the worst pain possible) and the Quick Disability of the Arm, Shoulder, and Hand questionnaire (QuickDASH, 0 = no disability, 100 = extreme disability) were completed by patients at the final follow-up. RESULTS: The average time of the procedure was 74 min. There were no perioperative complications. Average radiographic measures were: volar tilt, 8°; radial inclination, 23°, and ulnar variance, 0 mm. Wrist range of motion averaged 63° in wrist extension (range, 55-80°) and 55° in wrist flexion (range, 45-65°). Grip strength averaged 81% of the contralateral side at final evaluation (range, 67-100%). The mean QuickDASH score was 3.0 points (range, 0-9.1 points) and the mean VAS for pain was 9.3 at final follow-up. CONCLUSIONS: Open reduction internal fixation with the DePuy-Synthes 2.0 mm and 2.4 mm locking plates is an effective means of fixing a distal radius fracture that includes a volar rim fragment without interfering with flexor tendon gliding.
Subject(s)
Bone Plates , Fracture Fixation, Internal/instrumentation , Fracture Fixation, Internal/methods , Radius Fractures/surgery , Adult , Aged , Aged, 80 and over , Disability Evaluation , Female , Follow-Up Studies , Hand Strength , Humans , Male , Middle Aged , Open Fracture Reduction , Operative Time , Prospective Studies , Range of Motion, Articular , Visual Analog ScaleABSTRACT
BACKGROUND: Corrective surgery for kyphoscoliosis patients tend to be highly invasive due to osteotomy. The present case introduce less invasive corrective surgery using anterior oblique lateral interbody fusion (OLIF) technique. CASE PRESENTATION: An 80-year-old Japanese man with a history of Parkinson's disease presented to our hospital because of severe kyphoscoliosis and gait disturbance. Considering the postsurgical complications due to osteotomy, we performed an anterior-posterior combined corrective fusion surgery: OLIF of Lumbar (L) 2-3, L3-4, and L4-5 (Medtronic Sofamor Danek, Memphis, TN, USA) followed by L5-Sacral (S) 1 anterior lumbar fusion via the OLIF approach using an anterior intervertebral cage, and posterior L3-4 and L4-5 facetectomy and posterior fusion using percutaneous pedicle screws from Thoracic (T) 10 to S1 with a T-9 hook system. The surgery was performed in a less invasive manner with no osteotomy, and it improved the sagittal alignments with moderate restoration, which improved the patient's posture and gait disturbance. The patient showed transient muscle weakness of proximal lower extremity contralateral side to the surgical site, which fully recovered by physical rehabilitation 3 months after the surgery. CONCLUSION: The surgical corrective procedure using the minimally invasive OLIF method including L5-S1 fusion showed a great advantage in treating degenerative kyphoscoliosis in a Parkinson's disease patient in its less-invasive approac.