Introduction
The 2024 revisions of the McDonald criteria for multiple sclerosis (MS), published in Lancet Neurology in October 2025, formally incorporate the paramagnetic rim lesion (PRL) into the diagnostic framework.1 Under the revised criteria, in patients with a typical clinical presentation and typical lesions confined to a single central nervous system region, the presence of ≥ 1 PRL together with dissemination in time (DIT) or positive cerebrospinal fluid (CSF) (oligoclonal bands or kappa free light chain index) is sufficient to diagnose MS. This addition increases diagnostic specificity in specific situations.2 With this change, the detection of PRLs—chronic active lesions whose peripheral iron-laden microglia and macrophages produce a characteristic rim—has moved from a research setting into routine diagnostic practice. PRLs are primarily detected on phase images derived from susceptibility-weighted imaging (SWI) or other susceptibility-sensitive sequences rather than on conventional magnitude-based images; the phase signal reflects the paramagnetic susceptibility of iron within the rim, which is why appropriate phase-image display is central to their identification.
At the same time, the prognostic relevance of PRLs has continued to accumulate. Recent longitudinal data demonstrate that the evolution of PRLs—both the resolution of existing rims and the appearance of new ones—is associated with clinical worsening and disability progression and that baseline PRL count is increasingly being evaluated as a prognostic and treatment-response biomarker.3 Taken together, the revised criteria and these outcome findings reinforce an uncomfortable truth: PRL assessment must now be accurate, reproducible, and broadly implementable across centers if it is to fulfill its new diagnostic and prognostic role.
The accompanying 2024 MAGNIMS-CMSC-NAIMS consensus recommendations provide a much-needed framework for PRL acquisition, addressing field strength, voxel size, and flip angle.4 Complementing these recommendations, the 2024 NAIMS consensus document on chronic active lesion imaging provides essential histopathological correlation alongside practical interpretive guidance.5 However, neither document discusses the display settings applied at the reading workstation—in particular, the phase-image window/level. PRL detection requires careful attention not only to acquisition conditions6,7 but also to the conditions under which the images are reviewed. Even with optimal acquisition, an inappropriate phase window can render the paramagnetic rim invisible, leading to systematic undercounting of PRLs in patients who would otherwise meet the new criteria.
Practical Tips for Phase-image Review
1. Know your scanner’s phase convention
A paramagnetic rim may appear hypo- or hyperintense depending on the reconstruction convention, and there is no universally accepted convention for phase-image display.5 A practical rule is that the rim signal should always match that of veins, which contain paramagnetic deoxyhemoglobin.5 The displayed polarity is predetermined by the vendor implementation and is not ordinarily a user-selectable parameter; phase handedness varies across vendors and systems.8 Clinicians should therefore judge the rim in relation to the signal of veins rather than relying on a fixed “dark rim” or “bright rim” appearance.
2. Windowing is the most under-appreciated step
Phase images in routine MRI interpretation are most often viewed with vendor-default wide windows. For PRL detection, such windows commonly obscure both the rim and the central vein sign. In our experience, narrowing the window width first and then adjusting the window level substantially improves conspicuity; as a practical guide, narrowing the window until the cortex and white matter are clearly separated approximates the conditions under which PRLs are usually appreciated (Fig.
1). The stepwise application of this adjustment—first narrowing the window width and then adjusting the window level—is demonstrated in real time from Fig. 1c to 1d in the online Supplementary Fig.
S1. The optimal width currently relies on empirical experience rather than systematic validation, and formal reader studies comparing window settings would be valuable.
3. Hands-on training is desired
Recognizing PRLs depends on familiarity with the phase-image appearance of chronic active lesions on one’s own scanner and viewer, including how windowing affects conspicuity; in our experience, this skill is acquired through supervised review of examples rather than from textbooks alone. Trainees and clinicians should be exposed to worked examples across vendor conventions before reading susceptibility-sensitive images in clinical practice; this is particularly important under the 2024 criteria, in which a single overlooked PRL can determine whether the threshold for diagnosis (≥ 1 PRL plus DIT or positive CSF in the single-region scenario) is met.
Looking Ahead
As the 2024 McDonald criteria begin to be implemented worldwide, formal reader studies comparing acquisition and display settings, together with consensus language that explicitly covers the phase-image windowing step, will be needed if PRL assessment is to be reproducible across scanners and routine clinical environments. We hope that this brief practical editorial will help refocus attention on the under-recognized role of phase-image display in PRL detection—and, ultimately, on the patients whose diagnosis and prognosis depend on it.