Feasibility of Single and Multiplane Cine MR for Monitoring Tumor Volumes and Organs-at-Risk (OARs) Position During Radiation Therapy

Abstract

Purpose/Objective(s)Multiple groups are working on integrating MR guidance during radiation therapy. A systematic evaluation of images is required to identify targets and OARs that may be tracked or intervened on. The feasibility of target and OAR visualization with cine MR acquired using an integrated MR-Co-60 system was evaluated in this study.Materials/MethodsNineteen patients undergoing radiation therapy were enrolled on an IRB approved protocol, and underwent MR imaging on the 0.3T MR-Co-60 system. 35 cine image sets were acquired in the transverse (n = 10), sagittal (n = 18), or coronal (n = 7) plane with a slice thickness of 5 (n = 27) or 7 (n = 8) mm. The acquisition rate was 4 frames per second (fps) for images acquired in a single plane (n = 18), and 2 fps for images acquired simultaneously in 3 parallel planes (n = 17). OAR were selected from a published naming taxonomy. All radiation therapy targets and OARs that were within the imaging field of view for each dataset were evaluated by 3 radiation oncologists (JRO, PJP, CRS) to study the feasibility of monitoring anatomical motion. Physicians were specifically asked if image quality would permit manual visualization of the target or OAR for the purpose of gated radiation therapy.ResultsPhysicians evaluated 21 total target structures, and 321 total OAR structures in 35 image sets, including 13 unique targets and 58 unique OARs that were within the field of view. 7/13 targets were identified with consensus agreement by all 3 physicians that image quality was suitable for manual gating. Radiation therapy targets that were well visualized included tumors of the cervix, liver, lung, nasopharynx, and the breast lumpectomy cavity. 44/58 OARs were identified with consensus agreement by all physicians that image quality was suitable for manual gating. Well visualized OARs included the aorta, atrium, base of tongue, bladder, brainstem, breast, bronchial tree, carina, cerebellum, cerebrum, colon, duodenum, esophagus, femur, heart ventricle, kidney, larynx, liver, lung, mainstem bronchus, masseter muscle, optic nerve, pelvic bones, pelvic vessels, penile bulb, penis, pericardium, pharyngeal constrictors, pituitary, prostate, pulmonary vessels, rectum, retina (globe), rib, sacrum, seminal vesicles, small bowel, spinal cord, spleen, stomach, subclavian vessels, tongue, trachea, uterus, and vertebral body. A unanimous consensus was reached for 19/21 (90%) of targets and 308/321 (96%) of critical structures evaluated.ConclusionsA wide variety of radiation therapy targets and OARs were well visualized on single and multi-plane cine 0.3T MRI. Future work including testing automated tracking algorithms and developing clinical trials for the visualized targets and OARs is needed to fully exploit MRI-guided radiation therapy. Purpose/Objective(s)Multiple groups are working on integrating MR guidance during radiation therapy. A systematic evaluation of images is required to identify targets and OARs that may be tracked or intervened on. The feasibility of target and OAR visualization with cine MR acquired using an integrated MR-Co-60 system was evaluated in this study. Multiple groups are working on integrating MR guidance during radiation therapy. A systematic evaluation of images is required to identify targets and OARs that may be tracked or intervened on. The feasibility of target and OAR visualization with cine MR acquired using an integrated MR-Co-60 system was evaluated in this study. Materials/MethodsNineteen patients undergoing radiation therapy were enrolled on an IRB approved protocol, and underwent MR imaging on the 0.3T MR-Co-60 system. 35 cine image sets were acquired in the transverse (n = 10), sagittal (n = 18), or coronal (n = 7) plane with a slice thickness of 5 (n = 27) or 7 (n = 8) mm. The acquisition rate was 4 frames per second (fps) for images acquired in a single plane (n = 18), and 2 fps for images acquired simultaneously in 3 parallel planes (n = 17). OAR were selected from a published naming taxonomy. All radiation therapy targets and OARs that were within the imaging field of view for each dataset were evaluated by 3 radiation oncologists (JRO, PJP, CRS) to study the feasibility of monitoring anatomical motion. Physicians were specifically asked if image quality would permit manual visualization of the target or OAR for the purpose of gated radiation therapy. Nineteen patients undergoing radiation therapy were enrolled on an IRB approved protocol, and underwent MR imaging on the 0.3T MR-Co-60 system. 35 cine image sets were acquired in the transverse (n = 10), sagittal (n = 18), or coronal (n = 7) plane with a slice thickness of 5 (n = 27) or 7 (n = 8) mm. The acquisition rate was 4 frames per second (fps) for images acquired in a single plane (n = 18), and 2 fps for images acquired simultaneously in 3 parallel planes (n = 17). OAR were selected from a published naming taxonomy. All radiation therapy targets and OARs that were within the imaging field of view for each dataset were evaluated by 3 radiation oncologists (JRO, PJP, CRS) to study the feasibility of monitoring anatomical motion. Physicians were specifically asked if image quality would permit manual visualization of the target or OAR for the purpose of gated radiation therapy. ResultsPhysicians evaluated 21 total target structures, and 321 total OAR structures in 35 image sets, including 13 unique targets and 58 unique OARs that were within the field of view. 7/13 targets were identified with consensus agreement by all 3 physicians that image quality was suitable for manual gating. Radiation therapy targets that were well visualized included tumors of the cervix, liver, lung, nasopharynx, and the breast lumpectomy cavity. 44/58 OARs were identified with consensus agreement by all physicians that image quality was suitable for manual gating. Well visualized OARs included the aorta, atrium, base of tongue, bladder, brainstem, breast, bronchial tree, carina, cerebellum, cerebrum, colon, duodenum, esophagus, femur, heart ventricle, kidney, larynx, liver, lung, mainstem bronchus, masseter muscle, optic nerve, pelvic bones, pelvic vessels, penile bulb, penis, pericardium, pharyngeal constrictors, pituitary, prostate, pulmonary vessels, rectum, retina (globe), rib, sacrum, seminal vesicles, small bowel, spinal cord, spleen, stomach, subclavian vessels, tongue, trachea, uterus, and vertebral body. A unanimous consensus was reached for 19/21 (90%) of targets and 308/321 (96%) of critical structures evaluated. Physicians evaluated 21 total target structures, and 321 total OAR structures in 35 image sets, including 13 unique targets and 58 unique OARs that were within the field of view. 7/13 targets were identified with consensus agreement by all 3 physicians that image quality was suitable for manual gating. Radiation therapy targets that were well visualized included tumors of the cervix, liver, lung, nasopharynx, and the breast lumpectomy cavity. 44/58 OARs were identified with consensus agreement by all physicians that image quality was suitable for manual gating. Well visualized OARs included the aorta, atrium, base of tongue, bladder, brainstem, breast, bronchial tree, carina, cerebellum, cerebrum, colon, duodenum, esophagus, femur, heart ventricle, kidney, larynx, liver, lung, mainstem bronchus, masseter muscle, optic nerve, pelvic bones, pelvic vessels, penile bulb, penis, pericardium, pharyngeal constrictors, pituitary, prostate, pulmonary vessels, rectum, retina (globe), rib, sacrum, seminal vesicles, small bowel, spinal cord, spleen, stomach, subclavian vessels, tongue, trachea, uterus, and vertebral body. A unanimous consensus was reached for 19/21 (90%) of targets and 308/321 (96%) of critical structures evaluated. ConclusionsA wide variety of radiation therapy targets and OARs were well visualized on single and multi-plane cine 0.3T MRI. Future work including testing automated tracking algorithms and developing clinical trials for the visualized targets and OARs is needed to fully exploit MRI-guided radiation therapy. A wide variety of radiation therapy targets and OARs were well visualized on single and multi-plane cine 0.3T MRI. Future work including testing automated tracking algorithms and developing clinical trials for the visualized targets and OARs is needed to fully exploit MRI-guided radiation therapy.