Exploration of Digital Health Technologies

Table 1. Performance comparison of recent DARWIN-based studies on early Alzheimer’s detection.

Study	Dataset	Features/Models	Validation	Performance (accuracy/AUC)	Notes
Cilia et al., 2018 [16]	Darwin	450 features, dimensionality reduction	Train/test split	Accuracy (N/A)/AUC/(N/A)	Identified in-air time as key
Demircioglu Diren, 2025 [1]	Darwin	Dimensionality reduction + explainability techniques	CV	Accuracy 0.962/AUC(N/A)	High performance using feature selection
Kang et al., 2024 [18]	Darwin	Self-attention	CV	Accuracy 0.943/AUC(N/A)	Outperformed CNNs
Gong et al., 2025 [3]	Darwin	Hybrid transformer	CV	Accuracy 0.909/AUC(N/A)	Multimodal 2D + 1D features
Bazarbekov et al., 2026 [14]	Sensor-based Smart Pen dataset	Hybrid CNN-BiLSTM + Sim-to-Real Domain Adaptation	Train/test split	Accuracy 0.91/AUC 0.96	Deep learning, physics-based augmentation
Current study	Darwin	14 temporals + kinematic + pressure features, SVM/RF	5-fold CV	Accuracy 0.844/AUC 0.923	Interpretable, stroke-level temporal analysis

Return to article view

Metrics were extracted from the original publications. Missing values are indicated as N/A. All numerical values are rounded to three decimal places for consistency. Bazarbekov et al. (2026) [14] used a proprietary Smart Pen dataset with sensor-based motion acquisition rather than the publicly available DARWIN dataset.

Declarations

Acknowledgments

The author thanks Michael Mutsa Munanairi for assistance with proofreading and language editing.

Author contributions

CN: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing—original draft, Writing—review & editing. The author read and approved the final version of the manuscript.

Conflicts of interest

The author declares no conflicts of interest.

Ethical approval

This study uses the DARWIN dataset, an openly available public dataset released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Ethical approval for the original data collection was obtained by the investigators who created and published the DARWIN dataset. Because this study conducts secondary analysis of publicly available, de-identified data, no additional ethical approval was required.

Consent to participate

Informed consent for participation was obtained by the original investigators during the initial DARWIN data collection. This study did not involve any new data collection; the dataset was only downloaded and used in accordance with the dataset’s license and citation requirements. Therefore, no new participant consent was necessary.

Consent to publication

Consent for publication of the DARWIN dataset was obtained by the original investigators during the initial data collection process. The dataset is publicly distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license, which permits reuse, adaptation, and redistribution with appropriate attribution. Because this study involves only secondary analysis of the openly available, de-identified DARWIN dataset, no additional consent for publication was required.

Availability of data and materials

The datasets analyzed for this study can be found in the UCI Machine Learning Repository (link: https://archive.ics.uci.edu/datasets/?search=Darwin).

Funding

No funding was received for this study.

Copyright

Publisher’s note

Open Exploration maintains a neutral stance on jurisdictional claims in published institutional affiliations and maps. All opinions expressed in this article are the personal views of the author(s) and do not represent the stance of the editorial team or the publisher.

References

Demircioglu Diren D. Design and Validation of a Hybrid Machine Learning Model for Alzheimer’s Detection Using Handwriting Data. Am J Alzheimer’s Dis Other Dementias. 2025;40:15333175251374913. [DOI] [PubMed] [PMC]

Garre-Olmo J, Faúndez-Zanuy M, López-de-Ipiña K, Calvó-Perxas L, Turró-Garriga O. Kinematic and Pressure Features of Handwriting and Drawing: Preliminary Results Between Patients with Mild Cognitive Impairment, Alzheimer Disease and Healthy Controls. Curr Alzheimer Res. 2017;14:960–8. [DOI] [PubMed] [PMC]

Gong C, Qin H, El-Yacoubi MA. Hybrid Transformer for Early Alzheimer’s Detection: Integration of Handwriting-Based 2D Images and 1D Signal Features. IEEE J Biomed Health Inform. 2025;29:8293–305. [DOI] [PubMed]

Li S, Li K, Liu J, Huang S, Wang C, Tu Y, et al. A new method for community-based intelligent screening of early Alzheimer’s disease populations based on digital biomarkers of the writing process. Front Comput Neurosci. 2025;19:1564932. [DOI]

van den Berg RL, van der Landen SM, Keijzer MJ, van Gils AM, van Dam M, Ziesemer KA, et al. Smartphone- and Tablet-Based Tools to Assess Cognition in Individuals With Preclinical Alzheimer Disease and Mild Cognitive Impairment: Scoping Review. J Med Internet Res. 2025;27:e65297. [DOI] [PubMed] [PMC]

van den Berg RL, de Boer C, Zwan MD, Jutten RJ, van Liere M, van de Glind MABJ, et al. Digital remote assessment of speech acoustics in cognitively unimpaired adults: feasibility, reliability and associations with amyloid pathology. Alzheimers Res Ther. 2024;16:176. [DOI] [PubMed] [PMC]

Thebaud T, Favaro A, Chen C, Chavez G, Moro-Velazquez L, Butala A, et al. Explainable Metrics for the Assessment of Neurodegenerative Diseases through Handwriting Analysis. arXiv:2409.08303v2 [Preprint]. 2025 [cited 2025 Nov 6]. Available from: https://arxiv.org/html/2409.08303v2

Pinto TCC, Machado L, Bulgacov TM, Rodrigues-Júnior AL, Costa MLG, Ximenes RCC, et al. Is the Montreal Cognitive Assessment (MoCA) screening superior to the Mini-Mental State Examination (MMSE) in the detection of mild cognitive impairment (MCI) and Alzheimer’s Disease (AD) in the elderly? Int Psychogeriatr. 2019;31:491–504. [DOI] [PubMed]

Prince M, Wimo A, Guerchet M, Ali GC, Wu YT, Prina M. World Alzheimer report 2015. The global impact of dementia: an analysis of prevalence, incidence, costand trends. Doctoral dissertation. Alzheimer’s Disease International; 2015.

10.

Sperling RA, Aisen PS, Beckett LA, Bennett DA, Craft S, Fagan AM, et al. Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging‐Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7:280–92. [DOI] [PubMed] [PMC]

11.

Jack CR Jr, Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, et al. NIA-AA research framework: toward a biological definition of Alzheimer’s disease. Alzheimers Dement. 2018;14:535–62. [DOI] [PubMed] [PMC]

12.

Hosmer DW Jr, Lemeshow S, Sturdivant RX. Applied logistic regression. 3rd ed. Hoboken (NJ): John Wiley & Sons; 2013.

13.

Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396:413–46. [DOI] [PubMed] [PMC]

14.

Bazarbekov I, Almisreb A, Ipalakova M, Bazarbekova M, Daineko Y. Sim-to-Real Domain Adaptation for Early Alzheimer’s Detection from Handwriting Kinematics Using Hybrid Deep Learning. Sensors. 2026;26:298. [DOI] [PubMed] [PMC]

15.

Cilia ND, De Gregorio G, De Stefano C, Fontanella F, Marcelli A, Parziale A. Diagnosing Alzheimer’s disease from on-line handwriting: A novel dataset and performance benchmarking. Eng Appl Artif Intell. 2022;111:104822. [DOI]

16.

Cilia ND, Stefano CD, Fontanella F, Di Freca AS. An Experimental Protocol to Support Cognitive Impairment Diagnosis by using Handwriting Analysis. Procedia Comput Sci. 2018;141:466–71. [DOI]

17.

DARWIN dataset [Internet]. UCI Machine Learning Repository; [cited 2025 Nov 5]. Available from: https://archive.ics.uci.edu/datasets/?search=DARWIN

18.

Kang L, Zhang X, Guan J, Huang K, Wu R. Early Alzheimer’s disease diagnosis via handwriting with self-attention mechanisms. J Alzheimers Dis. 2024;102:173–80. [DOI] [PubMed]