Bridging the Gap: Lessons Learned from Radiation Oncology

MRI is increasingly being integrated into IGRT, driving a paradigm shift toward MRIgRT. The role of MRI in RT spans multiple stages: from initial dosimetry treatment plans via MR simulation to mid-treatment MRI for offline replanning in response to tumor changes, and using in-treatment MRI, as part of state-of-art MR Linear accelerator (MR-LINAC) system for real-time online adaptation. This enables dosimetry plans to be dynamically adjusted to account for changes in a patient’s anatomy and disease state during each treatment session. 

MRI simulation is performed on a dedicated MRI scanner, known as MR simulator, a whole-body MRI system with hardware and software comparable to diagnostic MRI but equipped with specialized patient positioning devices to replicate the treatment setup. Typically, a 1.5T scanner is used as MR simulator, though a 3.0T scanner may be employed when functional MRI is a clinical focus.  Imaging protocols for MR simulation are optimized for RT treatment planning, balancing factors such as the selection of pulse sequences, parameter optimization to achieve high SNR and spatial resolution while ensuring sufficient anatomical coverage and realistic acquisition time, and most importantly, minimizing geometric distortion. Additional considerations include coil configuration and respiratory gating strategies to accommodate immobilization devices and variations in patient body shape.

Two major commercial MR-LINAC systems are available for clinical use: the 0.35T MRIdian (ViewRay, USA) and the 1.5T Unity (Elekta, Sweden). These systems feature horizontal closed-bore MRI hardware specifically designed to allow radiation beams to pass through key MRI components. However, the MRI hardware in these MR-LINAC systems differs from conventional diagnostic MRI, introducing limitations such as B0 field inhomogeneity, reduced gradient performance, and lower SNR, all of which can impact image quality. Therefore, MR-LINAC imaging protocols must be optimized to account for these hardware and software constraints while still delivering sufficient image quality for onboard MRI-guided plan re-optimization and real-time motion monitoring during radiation delivery. Additionally, the imaging parameters should align as closely as possible with MRI simulation protocols to ensure that physicians can accurately assess changes in tumors and organs-at-risk (OARs) when comparing onboard images with reference images acquired on an MRI simulator.

Beyond understanding MRI techniques and optimizing imaging protocols for current clinical RT applications on MR simulators and MR-LIINACs, research in MR-in-RT is advancing in three key areas: (1) enhancing image quality and acquisition speed, (2) developing 4D MRI and real-time MRI for motion estimation and tracking, and (3) incorporating biological insights into the adaptive optimization process, referred to as biological image-guided adaptive radiotherapy, ultimately enabling personalized RT. These research areas are evolving in parallel, each requiring specialized expertise across scientific, technical, and clinical disciplines. The transition from technical validation and QA/QC program development to validation in clinical trials, and eventual integration into clinical practice remains a long-term endeavor.

There is an urgent need for close collaboration among three key clinical role pairs: MRI scientists/physicist and therapeutic physicist, MRI technologist and radiation therapist, and radiologist and radiation oncologist. Understanding both inter-pair and intra-pair gaps is essential for advancing clinical MRIgRT. A dedicated MR-in-RT physicist within the radiation oncology department plays a pivotal role in bridging these gaps. This individual must possess expertise in MRI physics and imaging technologies, clinical protocol optimization, QA/QC for both diagnostic and RT applications, as well as hardware and software utilization for clinical practice. Beyond technical responsibilities, the MR-in-RT physicist serves as a crucial liaison, facilitating communications among stakeholders. This includes educating radiation oncologists and therapists on both clinical applications and advanced imaging techniques, collaborating with imaging scientists and researchers to design and develop advanced acquisition and reconstruction methods for MRIgRT, and working with vendor and software developers to translate these innovations into clinical utility.

In summary,Summary:

· Technical challenges and optimization needs – MRI simulation and MR-LINAC imaging require careful protocol optimization to balance image quality, acquisition speed, and geometric accuracy, with specific constraints imposed by hardware limitations.

·  Interdisciplinary collaboration for clinical practice and innovation – Advancing MRIgRT necessitates close coordination among MRI scientists and physicists, therapeutic physicists, MRI technologists, radiation therapists, radiologists, and radiation oncologists to bridge knowledge gaps and optimize clinical workflows.

·  Future directions and research priorities – Key areas of ongoing research include enhancing MRI image quality and speed, developing real-time 4D MRI for motion tracking, and integrating biological insights into adaptive radiotherapy for truly personalized treatment.