Ge discovery 750

MRI examination was performed on a 3T scanner (GE Discovery 750, GE Healthcare, Milwaukee, WI, USA) using a dedicated MRI endometrial cancer staging protocol, which including sagittal, axial and coronal T2-weighted fast spin-echo, axial T1-weighted fast spin-echo, axial diffusion-weighted imaging echo planar imaging and sagittal dynamic contrast-enhanced T1-weighted fast spin-echo.

The performances of the prediction models were assessed using the receiver operating characteristic (ROC) curve and area under the curve (AUC) with a 95% confidence interval (CI). The sensitivity, specificity, and accuracy were also calculated for further evaluation. The calibration curves were applied to modify and reduce the bias of the models. Furthermore, the performance of the constructed models in this study were evaluated by a 5-fold cross-validation setup.
For comparison, a senior and junior head and neck radiologist (Y.Y. and JH.T.) were invited to independently diagnose IP-SCC via MR images. Both of the radiologists were blinded to the histopathology during the diagnosis process.
All of the images were scanned by three MR scanners (Philips Ingenia; GE Signa HDxt; GE Discovery 750). Stratified analysis was used to check whether the models were affected by different MR scanners.

All MRI examinations were performed with a 3.0 T MRI scanner (GE Discovery 750, GE Healthcare, Milwaukee, WI, USA) with an 8‑channel head coil. All patients received brain MRI examination including DWI, and FSE-Cube. The FSE-Cube sequence was obtained in a coronal plane with the following parameters: repetition time/echo time (TR/TE) 800/16 ms; slice thickness, 0.6 mm; slice gap, 0.3 mm; slice number, 256; image matrix 480 × 320; field of view (FOV) 23.0 cm × 18.4 cm; voxel size = 0.48 mm × 0.57 mm × 0.6 mm and scan duration, 6 min 32 s.
All patients underwent 3D DSA examination (Siemens Artist Zeego System, Siemens AG, Forchheim, Germany) while under local anesthesia. The tip of the catheter was positioned in the common carotid artery. Standard two-dimensional (2D) DSA was initially performed; 3D rotational images were then obtained from a series of subtracted images acquired with a rotation of a vascular C‑arm [19 (link)] and contrast injection (3 ml/s × 6 s, iopamidol 300 mg/ml) with the following parameters: angle 200º, angulation step 1.5º/f, and FOV 42 cm × 42 cm. The rotational angiographic data were reconstructed using dual volume model, then the mask volume and subtraction volume were obtained.

Imaging was performed using a 3.0 T whole-body MRI system (GE Discovery 750; GE Healthcare, Waukesha, Wisconsin, USA) with a 32-channel head coil. MRI protocols included 3D time-of-flight (TOF) magnetic resonance angiography, 2D high-resolution black-blood T2-weighted fast-spin-echo sequences, and pre- and post-contrast 3D high-resolution black-blood T1-weighted fast-spin-echo sequences, as well as conventional brain MRI. Enhanced T1WI was performed by obtaining repeated scans within 10 min after intravenous administration of gadopentetate dimeglumine injection (Beilu Pharmaceutical, Beijing, China) (0.1 mmol/kg, 2.0 mL/s). Detailed scan parameters are listed in Additional file 1: Table S1 Summery of imaging parameters.