AI and Neurological Rare & Neuroimmune Diseases ── Parkinson's ── 2026-10-28
Parkinson's disease and parkinsonism: treatment, trials and AI
- Abstract
BACKGROUND/OBJECTIVES: Machine learning (ML) is frequently proposed for Parkinson's disease (PD) diagnosis, but reported accuracy may not reflect methodological quality, generalizability, or clinical feasibility. We compared diagnostic performance, risk of bias, external validation, and deployment burden across biomarker modalities. METHODS: We searched PubMed/Medical Literature Analysis and Retrieval System Online (MEDLINE, Web of Science, and Institute of Electrical and Electronics Engineers (IEEE) Xplore for studies from January 2020 through March 2026 using a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020-compliant protocol registered with the International Platform of Registered Systematic Review and Meta-analysis Protocols (INPLASY). AI-assisted screening was audited in 147 records (Cohen's κ = 0.862). Reviewers verified extracted data and Risk Of Bias ASsessment Tool (PROBAST) judgments. Deployment burden was rated across five acquisition domains. Full-text assessment used a modality-stratified sample of retained records; we audited those not selected. RESULTS: Of 8135 deduplicated records, 2046 were retained after screening and 377 reports were sampled for full-text assessment; the final synthesis included 264 studies across nine modalities. Among 252 studies reporting accuracy, median best-model accuracy was 95.3% (range 64.9-100%). However, under our adapted PROBAST rules, 259 studies (98.1%) were at high overall risk of bias; 215 (81.4%) remained at high overall risk when the external-validation item was excluded, and only 20 (7.6%) met the external-validation definition. In ten within-study comparisons, external performance fell in eight. Median deployment burden ranged from 1.0 (voice/speech) to 4.2 (neuroimaging). CONCLUSIONS: Reported performance has advanced faster than evidence of generalizability and clinical implementation. Clinical translation will require subject-independent evaluation, external multisite validation, and transparent reporting of performance and deployment requirements. These proportions describe the sample assessed, not the entire PD machine learning literature.
Journal IF-equivalent: 4.3 (OpenAlex 2-year mean citedness, value as of 2026-10-26, retrieved 2026-10-28; not the official Clarivate IF)Reference: Pinto Neto O, Pinho TOR. Machine Learning for Parkinson’s Disease Biomarkers: A Systematic Review of Diagnostic Performance, Risk of Bias, and Clinical Translation Readiness. Brain Sci. 2026 Sep 26;16(10):1027. doi:10.3390/brainsci16101027.Checked: Abstract only - Abstract
Parkinson's disease (PD) is one of the most common neurodegenerative disorders in the aging population, which is commonly associated with motor abnormalities that can be reflected in hand-drawn geometric patterns. However, early detection remains challenging due to the lack of specific screening or tests, especially in rural areas with limited medical resources, leading to delays in medical intervention. To address these critical limitations, this paper proposes FusionNet, a dual-branch vision Transformer framework that combines spatial and frequency-domain information for handwriting-based PD screening. The spatial branch learns geometric and stroke-related representations directly from grayscale drawings, whereas the frequency branch processes log-scaled amplitude spectra obtained by two-dimensional fast Fourier transform. Features from the two branches are integrated using bidirectional cross-attention (BCA), allowing spatial and spectral representations to condition one another before classification. A staged training strategy was adopted in which the individual branches were trained separately before optimization of the fusion components. Experimental evaluations on the publicly available NewHandPD dataset were conducted across the three drawing tasks: hand-drawn circle, meander, and spiral images from 66 participants (31 individuals with PD and 35 healthy controls). Performance was evaluated using five-fold Group K-Fold cross-validation, demonstrating that FusionNet achieves the highest accuracy of 91.68% on the meander task and maintains >85% sensitivity with >88% specificity across all tasks. Because the study used a small single dataset, participant-level separation was not guaranteed for all tasks, and no independent external validation was performed, the reported results should be interpreted as preliminary and dataset-specific. Analysis of model behavior further indicates that spatial-domain predictions could be affected by paper texture, whereas the complementary frequency representation reduces this source of bias. Finally, the model was incorporated into a LINE-based prototype to demonstrate the technical feasibility of remote image submission and automated analysis. This implementation remains a proof of concept and requires external, prospective clinical validation before use in clinical decision-making.
Journal IF-equivalent: 5.3 (OpenAlex 2-year mean citedness, value as of 2026-10-06, retrieved 2026-10-07; not the official Clarivate IF)Reference: Lin TW, Li GJ, Lien WC, Peng YT, Lai YT, Chen YH. Dual-Branch Fusion Transformer Networks for Remote Parkinson’s Disease Screening via Hand-Drawn Graphics. Bioengineering. 2026 Sep 23;13(10):1101. doi:10.3390/bioengineering13101101.Checked: Abstract only