Virtual non-contrast enhanced magnetic resonance imaging (VNC-MRI)

TY - JOUR

T1 - Virtual non-contrast enhanced magnetic resonance imaging (VNC-MRI)

AU - Lindner, Thomas

AU - Debus, Hanna

AU - Fiehler, Jens

N1 - Copyright © 2021 Elsevier Inc. All rights reserved.

PY - 2021/9

Y1 - 2021/9

N2 - PURPOSE: Application of contrast agents (CA) is widely used in various clinical fields like oncology. Similar to approaches used in computed tomography, virtual non-contrast enhanced (VNC) images can be generated with the goal to supersede true non-contrast enhanced (TNC) images.METHODS: In MRI a T1-mapping sequence with variable flip angle (VFA) was used to acquire two images with different image contrast at the same time. To generate VNC images postprocessing based on this technique, an image-space based material decomposition algorithm was used. The inverse of a sensitivity matrix, consisting of intensity values for both VFA images and every material respectively, was used to determine the three material fractions and to calculate the final VNC images. The technique was tested on a 3 T scanner using a phantom and two in-vivo scans of patients with glioma and glioblastoma respectively. In all these cases the required six values were manually derived from the respective material or the background from both VFA images.RESULTS: Postprocessing results of the phantom show that the chosen materials can be separated and visualized individually and unwanted materials can be suppressed. In the VNC images of in-vivo scans the signal of the CA is removed successfully.CONCLUSION: It was shown that VNC images that match the visual impression of the TNC images can be generated, resulting in possibly reduced scan times and avoided mismatches due to movement of the patient.

AB - PURPOSE: Application of contrast agents (CA) is widely used in various clinical fields like oncology. Similar to approaches used in computed tomography, virtual non-contrast enhanced (VNC) images can be generated with the goal to supersede true non-contrast enhanced (TNC) images.METHODS: In MRI a T1-mapping sequence with variable flip angle (VFA) was used to acquire two images with different image contrast at the same time. To generate VNC images postprocessing based on this technique, an image-space based material decomposition algorithm was used. The inverse of a sensitivity matrix, consisting of intensity values for both VFA images and every material respectively, was used to determine the three material fractions and to calculate the final VNC images. The technique was tested on a 3 T scanner using a phantom and two in-vivo scans of patients with glioma and glioblastoma respectively. In all these cases the required six values were manually derived from the respective material or the background from both VFA images.RESULTS: Postprocessing results of the phantom show that the chosen materials can be separated and visualized individually and unwanted materials can be suppressed. In the VNC images of in-vivo scans the signal of the CA is removed successfully.CONCLUSION: It was shown that VNC images that match the visual impression of the TNC images can be generated, resulting in possibly reduced scan times and avoided mismatches due to movement of the patient.

KW - Algorithms

KW - Humans

KW - Magnetic Resonance Imaging

KW - Phantoms, Imaging

KW - Reproducibility of Results

KW - Tomography, X-Ray Computed

U2 - 10.1016/j.mri.2021.06.004

DO - 10.1016/j.mri.2021.06.004

M3 - SCORING: Journal article

C2 - 34119648

VL - 81

SP - 67

EP - 74

JO - MAGN RESON IMAGING

JF - MAGN RESON IMAGING

SN - 0730-725X

ER -