Objective: To assess the feasibility of using a novel cognitive and functional digital biomarker assessment, the Altoida NeuroMarker platform, to evaluate the presence of MCI in PD.
Background: In Parkinson’s disease (PD), early and precise detection of mild cognitive impairment (MCI) remains challenging inspite of therapeutic and prognostic implications. Traditional assessments (pen, paper or digitalized tests) rely on subjective evaluations with somewhat limited sensitivity to subtle cognitive changes. Machine learning (ML)-based digital cognitive assessments could improve early detection and monitoring of MCI.
Method: The Altoida NeuroMarker platform is a tablet-based assessment that evaluates cognitive and functional abilities through a 10-12 minutes battery of motoric, augmented reality (AR), and speech-based tasks capturing various cognitive and functional metrics. In this ongoing study, 18 patients have been studied and we present data from 11 all of whom were also assessed using traditional cognitive assessments .
Results: 11 patients (mean age 59.0 years (±11.2 (SD) range: 38–71), Hoehn-Yahr (HY) stage 2.91 (SD = 0.54, range: 2–4) had cognitive function score (Addenbrookes cognitive examination (ACE-III) mean = 88.0±8.50, minimental state examination (MMSE) mean = 26.2±3.56) and depressive symptoms (Beck Depression Inventory) mean 11.8±8.48. The Neuromarker correctly identified 9 participants with MCI on clinical testing. The MCI group had a higher mean age (60.6 ± 8.2, range: 38–71) compared to the non-MCI group (52.0±18.4, range: 39–65) but had similar HY stages. Cognitive performance, assessed via ACE-III, was lower in the MCI group (mean 87.0± 7.9, range: 70–96) compared to the non-MCI group (mean 97.0±2.83, range: 95–99). The MCI group also exhibited higher depressive symptoms, with a mean Beck Depression score of 13.0±9.2, range: 2–25) compared to 7.0±0.0, range: 7–7) in the non-MCI group.
Conclusion: Inspite of low numbers, this data for the first time in PD patients demonstrates the feasibility of using a machine learning-based digital cognitive assessment for sensitive detection of MCI and large scale studies are required.
To cite this abstract in AMA style:
K. Poplawska-Domaszewicz, A. Brek, K. Wu, N. Griffin, S. Fallone, M. Iulita, E. Streel, A. Kurta-Nowicka, K. Naskret-Gasik, J. Konczak, K. Ray Chaudhuri. Pilot Feasibility Data using Augmented Reality and Machine Learning-Based Digital Cognitive Assessment in Parkinson’s disease [abstract]. Mov Disord. 2025; 40 (suppl 1). https://www.mdsabstracts.org/abstract/pilot-feasibility-data-using-augmented-reality-and-machine-learning-based-digital-cognitive-assessment-in-parkinsons-disease/. Accessed November 20, 2025.« Back to 2025 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/pilot-feasibility-data-using-augmented-reality-and-machine-learning-based-digital-cognitive-assessment-in-parkinsons-disease/