- Deep Transfer Learning and Radiomics Feature Prediction of Survival of Patients with High-Grade Gliomas
Fifty patients with high-grade gliomas from the authors’ hospital and 128 patients with high-grade gliomas from The Cancer Genome Atlas were included in this study. For each patient, the authors calculated 348 hand-crafted radiomics features and 8192 deep features generated by a pretrained convolutional neural network. They then applied feature selection and Elastic Net-Cox modeling to differentiate patients into long- and short-term survivors. In the 50 patients with high-grade gliomas from their institution, the combined feature analysis framework classified the patients into long- and short-term survivor groups with a log-rank test P value <.001. In the 128 patients from The Cancer Genome Atlas, the framework classified patients into long- and short-term survivors with a log-rank test P value of .014. In conclusion, the authors report successful production and initial validation of a deep transfer learning model combining radiomics and deep features to predict overall survival of patients with glioblastoma from postcontrast T1-weighed brain MR imaging.
- Prediction of Hemorrhage after Successful Recanalization in Patients with Acute Ischemic Stroke: Improved Risk Stratification Using Dual-Energy CT Parenchymal Iodine Concentration Ratio Relative to the Superior Sagittal Sinus
The authors evaluated whether, in acute ischemic stroke, iodine concentration within contrast-stained parenchyma compared with an internal reference in the superior sagittal sinus on dual-energy CT could predict subsequent intracerebral hemorrhage in 71 patients. Forty-three of 71 patients had parenchymal hyperdensity on initial dual-energy CT. The median relative iodine concentration compared with the superior sagittal sinus was significantly higher in those with subsequent intracerebral hemorrhage (137.9% versus 109.2%). They conclude that in dual-energy CT performed within 1 hour following thrombectomy that the relative iodine concentration within contrast-stained brain parenchyma compared with that in the superior sagittal sinus was a more reliable predictor of ICH compared with the absolute maximum iodine concentration.
- Number Needed to Treat with Vertebral Augmentation to Save a Life
The purpose of this study was to calculate the number needed to treat to save 1 life at 1 year and up to 5 years after vertebral augmentation. A 10-year sample of the 100% US Medicare data base was used to identify patients with vertebral compression fractures treated with nonsurgical management, balloon kyphoplasty, and vertebroplasty. The number needed to treat was calculated between augmentation and nonsurgical management groups from years 1–5 following a vertebral compression fracture diagnosis, using survival probabilities for each management approach. The adjusted number needed to treat to save 1 life for nonsurgical management versus kyphoplasty ranged from 14.8 at year 1 to 11.9 at year 5. The adjusted number needed to treat for nonsurgical management versus vertebroplasty ranged from 22.8 at year 1 to 23.8 at year 5. The authors conclude that the NNT analysis of more than 2 million patients with VCF reveals that only 15 patients need to be treated to save 1 life at 1 year. This has an obvious clinically significant impact and given that all augmentation clinical trials are underpowered to detect a mortality benefit, this large dataset analysis reveals that vertebral augmentation provides a significant mortality benefit over nonsurgical management with a low NNT.