MRI Physics for Radiation Therapy Planning

Magnetic resonance imaging has become a central source of anatomical information in modern radiation oncology. Its excellent soft-tissue contrast can make a visible difference when clinicians define a prostate, brain lesion, cervical tumour, or head-and-neck target. The physics of magnetic resonance imaging for radiation therapy planning, however, extends well beyond producing sharp diagnostic pictures. Safe use depends on understanding magnetic fields, spatial encoding, geometric accuracy, motion, image registration, and the interaction between MRI data and treatment-dose calculations.

For Australian services, the subject is especially relevant as metropolitan cancer centres adopt hybrid MRI-guided systems while regional hospitals often rely on partnerships with larger networks. A planning image must fit the realities of Sydney, Melbourne, Brisbane, Perth, Adelaide, and remote areas alike: different equipment, varying staff availability, long patient travel times, and procurement decisions shaped by both public and private healthcare markets.

Why MRI Adds Value to Treatment Planning

Computed tomography remains the conventional foundation for radiation therapy planning because it provides electron-density information for dose calculation and shows the patient’s external contour. MRI contributes a different strength: tissue discrimination. Tumour boundaries, nerves, muscles, vessels, marrow, and treatment-sensitive organs can be easier to distinguish on T2-weighted, diffusion-weighted, or contrast-enhanced sequences.

The most useful role is often complementary rather than substitutive. A CT simulation scan may supply the attenuation map, while a registered MRI improves contouring. In an MRI-only workflow, synthetic CT or deformable image conversion is required to estimate tissue density. Each approach introduces assumptions that must be tested before it becomes part of routine clinical practice.

MRI can also support adaptive radiation therapy. Repeated images show anatomical changes during a treatment course, such as bladder filling, rectal deformation, weight loss, or tumour shrinkage. Online adaptation can then modify contours and beam delivery, although the process requires rapid quality assurance and close coordination between radiation therapists, physicists, radiation oncologists, and radiologists.

The Magnetic Resonance Signal and Spatial Information

MRI begins with hydrogen nuclei, chiefly in water and fat, behaving like tiny magnetic moments. A strong static field aligns a small excess of these moments. A radiofrequency pulse disturbs that alignment, and the nuclei then return towards equilibrium while emitting a measurable signal. T1 relaxation describes recovery along the main field, while T2 relaxation describes loss of coherent transverse magnetisation.

Scanner hardware converts this signal into spatial information. Gradient coils vary the magnetic field across the patient, allowing frequency and phase encoding. The measured data fill k-space, and mathematical reconstruction converts those samples into an image. Field strength, gradient performance, radiofrequency coils, pulse sequence design, voxel size, and acquisition time all influence the final result.

For planning, image quality must be judged against geometric fidelity. A sequence with excellent contrast can still be unsuitable if distortion shifts an organ by several millimetres. Susceptibility differences at air-tissue interfaces, dental materials, surgical clips, and implants may cause local warping. Gradient non-linearity and magnetic-field inhomogeneity create additional spatial errors, particularly away from the scanner’s isocentre.

Geometry, Distortion, and Dose Calculation

Radiotherapy margins can be smaller than the errors tolerated in routine diagnostic imaging. MRI quality assurance therefore includes geometric phantom tests, assessment across the intended field of view, and checks after hardware or software changes. A service should measure the distortion relevant to its planning region rather than relying only on a manufacturer’s general specification.

Registration between MRI and CT has its own uncertainty. Differences in patient position, immobilisation, bladder volume, rectal filling, breathing phase, and couch indexing can produce apparent anatomical changes. Rigid registration may be appropriate for a relatively stable brain, while pelvic planning may require carefully reviewed deformable registration. The registration method should never replace clinical examination of the fused images.

MRI signal intensity cannot be translated directly into electron density. If MRI is used as the primary planning image, a synthetic CT model must estimate stopping power from anatomy, segmentation, atlas information, or machine-learning methods. Metallic implants, unusual anatomy, and gas pockets can challenge these models. Independent dose checks and a clearly documented fallback pathway remain essential.

Building a Safe Australian Workflow

Australian departments operate under state and territory radiation legislation, while national guidance from the Australian Radiation Protection and Nuclear Safety Agency helps shape safe practice. MRI does not use ionising radiation, yet it involves serious projectile, implant, heating, acoustic-noise, and emergency risks. Local MRI screening, access control, emergency procedures, and device verification must sit alongside the radiation therapy quality system.

The Australian market also has practical constraints. A large public hospital in Melbourne or Sydney may have access to specialist physicists and a dedicated MRI-linac, whereas a regional service may send patients to a tertiary centre and receive images through a shared electronic record. Procurement should therefore consider service contracts, staff training, interoperability, downtime support, and the cost of maintaining a second planning pathway.

Useful implementation priorities include:

• Define the clinical sites and treatment techniques that genuinely benefit from MRI.
• Standardise immobilisation, patient positioning, coils, sequences, and scan geometry.
• Establish MRI safety screening for implants, devices, tattoos, monitoring equipment, and emergency access.
• Validate registration, synthetic CT, contour propagation, and dose calculation with independent tests.
• Record image-quality, distortion, and workflow incidents in a shared quality-improvement process.

Everyday practice matters as well. Australian patients may travel several hours from a regional town for simulation, so repeat imaging caused by an avoidable setup error has a real human and financial cost. Scheduling should allow time for screening, changing, positioning, coils, and motion coaching rather than treating MRI as a quick add-on to a CT appointment.

From Scanner Room to Treatment Unit

An effective pathway starts before acquisition. The team should decide whether the MRI is diagnostic, planning-specific, or both; select sequences that answer the contouring question; and reproduce the treatment position as closely as possible. Flat tabletops, compatible immobilisation devices, indexing, and external lasers help reduce transfer errors. When a diagnostic MRI is used, the limitations of a non-treatment position must be explicit.

Motion management is important in the thorax, upper abdomen, and pelvis. Respiratory triggering, breath-hold techniques, cine imaging, or four-dimensional approaches may provide useful information, but each has timing and reconstruction limitations. A fast sequence with modest contrast may be more valuable than a slow, high-resolution sequence if it captures anatomy at the correct respiratory state.

The final plan should retain traceability. Staff need to know which scan was used, how it was registered, whether anatomy changed between imaging and treatment, and who approved any contour edits. For MRI-guided online treatment, the process becomes a repeated loop of imaging, segmentation, planning, verification, and delivery. Automation can shorten that loop, but it cannot remove the need for trained human review.

Professional Learning and Practical Exchange

The archived 2014 Annual Scientific Meeting of the Canadian Organization of Medical Physicists illustrates how professional meetings connect technical science with operational detail. Sessions on imaging, planning, examinations, business, and clinical practice provide a useful model for how medical physicists share methods across institutions. The Banff setting also highlights the value of informal discussion, where a problem with registration or quality assurance can be compared with another centre’s experience.

Conference logistics influence participation more than they may appear to. Delegates planning attendance could review the accommodation details alongside the scientific programme, just as Australian teams must account for travel between campuses, on-call coverage, and accommodation near major cancer centres during training visits. These practical decisions affect who can attend and which skills return to the department.

Poster sessions are especially relevant to MRI-guided planning because early results often appear there before a method becomes standard practice. A concise report may reveal distortion measurements, synthetic CT validation, patient-positioning tests, or workflow timings that are directly transferable. Guidance on poster sessions reflects a wider professional habit: ask precise questions, examine methods closely, and distinguish promising research from validated clinical performance.

MRI planning is now a multidisciplinary field. The strongest services bring together imaging specialists, radiation oncologists, therapists, engineers, IT teams, and physicists. They also engage patients, whose comfort, ability to remain still, and understanding of repeated imaging can determine whether the technical protocol succeeds in practice.

A department preparing to expand MRI-based planning should begin with a documented clinical use case, then test the entire pathway from referral to treatment delivery. Measure geometric accuracy, registration uncertainty, scan time, repeat rates, staff workload, and patient experience. Use those results to refine protocols before increasing case numbers, and maintain a visible record of approvals, audits, and corrective actions. Building that evidence base will help Australian services turn magnetic resonance data into reliable, patient-centred radiation therapy.