Can a Deep-learning Model for the Automated Detection of Vertebral Fractures Approach the Performance Level of Human Subspecialists?

BACKGROUND: Vertebral fractures are the most common osteoporotic fractures in older individuals. Recent studies suggest that the performance of artificial intelligence is equal to humans in detecting osteoporotic fractures, such as fractures of the hip, distal radius, and proximal humerus. However, whether artificial intelligence performs as well in the detection of vertebral fractures on plain lateral spine radiographs has not yet been reported. QUESTIONS/PURPOSES: (1) What is the accuracy, sensitivity, specificity, and interobserver reliability (kappa value) of an artificial intelligence model in detecting vertebral fractures, based on Genant fracture grades, using plain lateral spine radiographs compared with values obtained by human observers? (2) Do patients' clinical data, including the anatomic location of the fracture (thoracic or lumbar spine), T-score on dual-energy x-ray absorptiometry, or fracture grade severity, affect the performance of an artificial intelligence model? (3) How does the artificial intelligence model perform on external validation? METHODS: Between 2016 and 2018, 1019 patients older than 60 years were treated for vertebral fractures in our institution. Seventy-eight patients were excluded because of missing CT or MRI scans (24% [19]), poor image quality in plain lateral radiographs of spines (54% [42]), multiple myeloma (5% [4]), and prior spine instrumentation (17% [13]). The plain lateral radiographs of 941 patients (one radiograph per person), with a mean age of 76 ± 12 years, and 1101 vertebral fractures between T7 and L5 were retrospectively evaluated for training (n = 565), validating (n = 188), and testing (n = 188) of an artificial intelligence deep-learning model. The gold standard for diagnosis (ground truth) of a vertebral fracture is the interpretation of the CT or MRI reports by a spine surgeon and a radiologist independently. If there were any disagreements between human observers, the corresponding CT or MRI images would be rechecked by them together to reach a consensus. For the Genant classification, the injured vertebral body height was measured in the anterior, middle, and posterior third. Fractures were classified as Grade 1 (< 25%), Grade 2 (26% to 40%), or Grade 3 (> 40%). The framework of the artificial intelligence deep-learning model included object detection, data preprocessing of radiographs, and classification to detect vertebral fractures. Approximately 90 seconds was needed to complete the procedure and obtain the artificial intelligence model results when applied clinically. The accuracy, sensitivity, specificity, interobserver reliability (kappa value), receiver operating characteristic curve, and area under the curve (AUC) were analyzed. The bootstrapping method was applied to our testing dataset and external validation dataset. The accuracy, sensitivity, and specificity were used to investigate whether fracture anatomic location or T-score in dual-energy x-ray absorptiometry report affected the performance of the artificial intelligence model. The receiver operating characteristic curve and AUC were used to investigate the relationship between the performance of the artificial intelligence model and fracture grade. External validation with a similar age population and plain lateral radiographs from another medical institute was also performed to investigate the performance of the artificial intelligence model. RESULTS: The artificial intelligence model with ensemble method demonstrated excellent accuracy (93% [773 of 830] of vertebrae), sensitivity (91% [129 of 141]), and specificity (93% [644 of 689]) for detecting vertebral fractures of the lumbar spine. The interobserver reliability (kappa value) of the artificial intelligence performance and human observers for thoracic and lumbar vertebrae were 0.72 (95% CI 0.65 to 0.80; p < 0.001) and 0.77 (95% CI 0.72 to 0.83; p < 0.001), respectively. The AUCs for Grades 1, 2, and 3 vertebral fractures were 0.919, 0.989, and 0.990, respectively. The artificial intelligence model with ensemble method demonstrated poorer performance for discriminating normal osteoporotic lumbar vertebrae, with a specificity of 91% (260 of 285) compared with nonosteoporotic lumbar vertebrae, with a specificity of 95% (222 of 234). There was a higher sensitivity 97% (60 of 62) for detecting osteoporotic (dual-energy x-ray absorptiometry T-score ≤ -2.5) lumbar vertebral fractures, implying easier detection, than for nonosteoporotic vertebral fractures (83% [39 of 47]). The artificial intelligence model also demonstrated better detection of lumbar vertebral fractures compared with detection of thoracic vertebral fractures based on the external dataset using various radiographic techniques. Based on the dataset for external validation, the overall accuracy, sensitivity, and specificity on bootstrapping method were 89%, 83%, and 95%, respectively. CONCLUSION: The artificial intelligence model detected vertebral fractures on plain lateral radiographs with high accuracy, sensitivity, and specificity, especially for osteoporotic lumbar vertebral fractures (Genant Grades 2 and 3). The rapid reporting of results using this artificial intelligence model may improve the efficiency of diagnosing vertebral fractures. The testing model is available at http://140.113.114.104/vght_demo/corr/. One or multiple plain lateral radiographs of the spine in the Digital Imaging and Communications in Medicine format can be uploaded to see the performance of the artificial intelligence model. LEVEL OF EVIDENCE: Level II, diagnostic study.

Can imaging distinguish between low-grade and dedifferentiated parosteal osteosarcoma?

BACKGROUND: Most instances of the parosteal osteosarcoma (OGS) are low-grade tumors. However, some parosteal OGSs undergo dedifferentiated transformation. Dedifferentiated parosteal OGS can cause distant metastasis and poor survival, and preoperative chemotherapy may be warranted. This study provides imaging clues for dedifferentiated parosteal OGS before treatment. METHODS: The study retrospectively enrolled 23 patients with histologically proven parosteal OGS, including 69.6% (n = 16) low-grade and 30.4% (n = 7) dedifferentiated types. Preoperative images including radiography and magnetic resonance imaging were reviewed. The following imaging parameters and clinical outcomes were evaluated: 1) average age; 2) sex; 3) tumor size; 4) presence of string sign; 5) necrosis; 6) hemorrhage; 7) solid soft tissue component; 8) perforating vessels; 9) ossification grade; 10) marginal ossification; 11) periosteal reaction; 12) sunburst reaction; 13) bone marrow edema; 14) bone marrow invasion; 15) perifocal soft tissue edema; 16) adjacent joint involvement; 17) adjacent neurovascular bundle compression; 18) regional lymph node; 19) bone metastasis; 20) preoperative lung metastasis; 21) follow-up lung metastasis; and 22) recurrence. RESULTS: The average maximal tumor sizes were 7.1 cm and 10.9 cm in low-grade and dedifferentiated types, respectively (p = 0.033). Sunburst periosteal reaction was visualized in two cases of low-grade type (12.5%) and four cases of the dedifferentiated type (57.1%) (p = 0.025) of parosteal OGS. None of our studied cases revealed preoperative lung metastasis. In the follow-up chest computed tomography, lung metastasis was noted in two cases of conventional type (14.2%), and four cases of dedifferentiated type (57.1%) (p = 0.040) of parosteal OGS. In receiver operating characteristic (ROC) curve analysis, the average tumor size and sunburst periosteal reaction showed good specificity (AUC = 0.070 and 0.072, respectively). CONCLUSION: Compared with low-grade types, dedifferentiated parosteal OGS exhibits a considerably larger tumor size, more sunburst periosteal reaction, and a more frequent development of lung metastasis in the disease course. Tumor size and sunburst periosteal reaction are the most crucial imaging diagnostic factors.

Imaging and pathological correlation of soft-tissue chondroma: a serial five-case study and literature review.

The imaging findings (X-ray and MRI) and patterns of calcification of five patients with pathologically proven soft-tissue chondroma were correlated with histopathology. The size ranged from 0.5 to 3 cm. Four showed calcifications: curvilinear, punctuate, mixed curvilinear, and punctuate patterns, and the other one with a dystrophic or homogenous dense pattern showed hypointensity on T1- and T2-weighted MR imaging. Histopathology showed hyaline cartilage with nests of chondrocytes in the lacunae. Soft-tissue chondroma is a rare, benign soft-tissue tumor. It should be differentiated from other soft-tissue masses, especially malignancy.