Automated analysis of CT-derived body composition parameters, especially the visceral fat index, can predict the new-onset risk for type 2 diabetes better than traditional anthropometric and clinical risk models.
