New Technology May Help Differentiate Between Parkinsonisms

A machine learning–based software tool (neuropacs; Gainesville, FL) analyzing 3T diffusion MRI scans was found to be effective in distinguishing between Parkinson disease (PD), multiple system atrophy (MSA), and progressive supranuclear palsy (PSP) in people with diagnosed movement disorders. Results of a prospective cohort study published in JAMA Neurology demonstrate the potential of Automated Imaging Differentiation for Parkinsonism (ADIP) for use in the diagnostic workup.

The study included 249 participants with PD, MSA, and PSP from 21 Parkinson Study Group centers in the United States and Canada from 2021 to 2024, along with a retrospective cohort of 396 patients. Diagnoses were confirmed by unanimous agreement among 3 movement disorder specialists or by postmortem pathology. The primary endpoint of the study was the ability for ADIP to differentiate between PD vs atypical parkinsonism, MSA vs PSP, PD vs MSA, and PD vs PSP, as measured by area under the receiver operating characteristic curve (AUROC).

In terms of results, the AIDP model was found to be robust in differentiating between:

• PD vs atypical parkinsonism (AUROC, .96; 95% CI, .93 to .99)
• MSA vs PSP (AUROC, .98; 95% CI, .96 to 1.00)
• PD vs MSA (AUROC, .98; 95% CI, .96 to 1.00)
• And PD vs PSP (AUROC, .98; 95% CI, .96 to 1.00)

According to the study authors, the ease of obtaining diffusion MRI images in comparison to dopamine transporter imaging (DaT SPECT), which requires the use of a radioactive drug and has high costs, positions ADIP as a potential tool to be included in the diagnostic workup of movement disorders, especially when combined with skin biopsy or synuclein seed aggregation assay (SAA) testing.

Kristophe Diaz, PhD, Executive Director and Chief Science Officer of CurePSP, commented on the results of the study by saying, "Proper diagnosis of PSP remains a critical need to decrease burden for people living with the disease and ensure they receive proper clinical care as soon as possible. The results of this study are exciting progress towards a non-invasive and clinically scalable solution to help improve diagnosis.”