Superior dorsal nigral hyperintensity depiction at 7 T MRI using CLEAR-DESS improves diagnosis performance of Parkinson’s disease

npj Parkinson's Disease (Nature) · 2026-05-28 · open original →

Reader summary

by xavier.grehant on 2026-05-30

MRI neuroimaging Iron Iron & ferroptosis

Deep inside the brain is a small region called the substantia nigra — Latin for "black substance" — whose dark colour comes from a pigment called neuromelanin. In Parkinson's disease (PD), the neurons that produce neuromelanin and dopamine gradually die, and that darkness fades. On a powerful MRI scanner set to the right mode, healthy people show a characteristic bright patch (called dorsal nigral hyperintensity, or DNH) in the back part of this region; in PD patients the patch is dim or absent. Detecting that difference is one of the most direct ways to image the hallmark damage of PD. This prospective study from Xiangya Hospital in China compared two ways of capturing DNH on a 7-tesla (7T) MRI — a very high-field research scanner roughly twice as strong as the most common hospital MRIs. The standard approach, called SWI (susceptibility-weighted imaging, which uses iron contrast), was pitted against a newer sequence called CLEAR-DESS, which is engineered to suppress confusing signals from fluid and blood vessels so the substantia nigra stands out more clearly.

The study enrolled 33 people with diagnosed PD and 52 healthy volunteers, all scanned with both sequences. Four independent radiologists, reading images without knowing who had PD, rated CLEAR-DESS dramatically better. At the level of individual brain structures, CLEAR-DESS reached sensitivity of 85–96% and specificity of 88–94% (meaning it correctly identified most PD cases and correctly cleared most healthy controls). SWI reached only 65–70% sensitivity and 74–89% specificity. At the whole-patient level the gap was equally large: CLEAR-DESS specificity ran as high as 100% versus 60–79% for SWI. All image-quality metrics — signal-to-noise and contrast-to-noise ratios — also favoured CLEAR-DESS (all p < 0.001). The study is small and was done at a single centre, so results need validation in larger, independent cohorts and in settings where 7T MRI is less tightly controlled.

For people living with PD or awaiting a diagnosis, the practical message is twofold. First, the choice of MRI scan sequence matters as much as the strength of the magnet: two scans on the same 7T machine can give very different answers. Second, 7T MRI scanners remain largely confined to research hospitals — they are not a routine clinical tool yet. If this finding is confirmed in larger studies, it strengthens the case for investing in 7T infrastructure and for using CLEAR-DESS rather than SWI when 7T is available. A clinician ordering a brain MRI today at a standard 1.5T or 3T centre cannot yet request CLEAR-DESS; but for patients enrolled in clinical trials or seen at specialist centres with 7T access, asking whether the most diagnostic sequence is being used is a reasonable question. Wider clinical adoption, if it comes, is probably five or more years away.

What this article adds

MRI neuroimaging (topic pending review): A prospective study of 33 PD patients and 52 healthy controls at 7T MRI found that the CLEAR-DESS sequence detected the loss of dorsal nigral hyperintensity with sensitivity of 85–96% and specificity of 88–94%, substantially outperforming the standard SWI sequence (sensitivity 65–70%, specificity 74–89%). At the whole-patient level, CLEAR-DESS specificity reached up to 100% vs 60–79% for SWI. The study establishes CLEAR-DESS as a candidate preferred sequence for 7T MRI-based PD diagnosis.
Iron: SWI — an iron-sensitive MRI sequence widely used for nigral iron mapping in PD — was directly compared against the iron-suppressing CLEAR-DESS sequence in 7T MRI. CLEAR-DESS, which attenuates both fluid and blood-vessel signal rather than exploiting iron contrast, significantly outperformed SWI on every diagnostic metric, suggesting that free-water and vessel suppression adds more diagnostic value than iron weighting alone for visualising the dorsal substantia nigra.
Iron & ferroptosis: The head-to-head comparison between iron-sensitive SWI and the iron-suppressing CLEAR-DESS sequence at 7T MRI provides indirect in-vivo evidence that iron accumulation in the substantia nigra, while detectable, is not the optimal contrast mechanism for diagnosing PD: neuromelanin/free-water contrast captured by CLEAR-DESS gave markedly higher diagnostic accuracy.

Reader summary by xavier.grehant

by xavier.grehant on 2026-05-30

MRI nigral imaging

The substantia nigra — the brain region whose gradual loss of dopamine-producing cells defines Parkinson's disease — contains a small structure called nigrosome-1 that shows up as a bright patch on certain MRI scans. Researchers call this bright patch dorsal nigral hyperintensity (DNH). In healthy brains it is clearly visible; in Parkinson's it tends to disappear as cells die. At ultra-high-field 7 Tesla (7T) MRI (roughly twice the magnetic strength of most hospital scanners), this feature can potentially be used as a diagnostic sign. This prospective study — involving 33 people with Parkinson's and 52 healthy volunteers, all scanned at 7T — compared two different MRI scanning methods for detecting DNH: the established susceptibility-weighted imaging (SWI) technique, which is sensitive to iron deposits, and a newer sequence called CLEAR-DESS (combined liquid and blood vessel attenuation with refined double echo steady state), purpose-built to make the DNH stand out more sharply.

The results strongly favored CLEAR-DESS. When four radiologists assessed scans independently, CLEAR-DESS correctly identified Parkinson's patients with a sensitivity of 91–94% and a specificity of 88–100% (meaning it correctly cleared healthy volunteers nearly every time). SWI reached only 80–89% sensitivity and 60–79% specificity. Both the image quality scores and objective signal measurements were significantly better with CLEAR-DESS (all p < 0.001). Importantly, different readers agreed closely with each other on CLEAR-DESS images — suggesting the advantage doesn't depend on unusual radiologist expertise.

For someone living with Parkinson's or a family member, what matters here is the broader direction: researchers are building MRI-based tools that could one day confirm a Parkinson's diagnosis with a standard brain scan rather than relying purely on clinical observation, which today takes an average of one to two years and still carries a meaningful error rate. This study is small and 7T MRI remains rare — mostly confined to research hospitals. CLEAR-DESS is not yet available as a clinical service anywhere. But it points toward a future where a high-field brain MRI could give a neurologist far more objective evidence at the time of first suspicion. Larger validation studies in broader populations, and wider deployment of 7T scanners, are the next necessary steps before this approach could reach routine care.

What this article adds

MRI nigral imaging (topic pending review): This prospective study of 85 participants (33 PD, 52 healthy controls) shows that a new 7T MRI sequence called CLEAR-DESS detects the loss of dorsal nigral hyperintensity — the bright nigrosome-1 patch in the substantia nigra — with substantially higher accuracy (patient-level AUC 0.91–0.96, sensitivity 91–94%, specificity 88–100%) than the currently used susceptibility-weighted imaging (AUC 0.79–0.87, sensitivity 80–89%, specificity 60–79%), establishing CLEAR-DESS as the leading candidate sequence for this diagnostic approach at ultra-high-field MRI.

Contribute an improved analysis →