Describir: Synthetic temporal bone CT generation from UTE-MRI using a cycleGAN-based deep learning model: advancing beyond CT-MR imaging fusion.