Document Type
Article
Publication Date
2021
Abstract
Background: We built a web-based application of the Archimedes spiral exam that implements clinically validated spiral metrics and tested drawing instructions to define a clinical workflow.
Methods: We designed an HTML5 and Javascript implementation of the spiral exam to run on mobile touchscreen devices. We then recruited 10 volunteers each for 2 experiments designed to validate the programmed spiral metrics and assess how instructions or drawing implement affect the results. In task one, volunteers drew 5 spirals each while following 6 different instruction sets (n=30 spirals each, n=300 spirals total) that varied by support of the drawing hand and tracing condition (either tracing a spiral template, drawing in-between it, or freehand). In task two, volunteers drew 5 spirals each while following 2 instruction sets and drawing using a stylus or their dominant index finger (n=20 spirals each, n=200 spirals total).
Results: Principal components analysis of calculated metrics revealed that the experiments grouped by instruction set and by subject. Mean Euclidean distance between experiments represented as 11-dimensional vectors revealed that consistency varied among instruction tasks and that drawing with a stylus produced more consistent results than did using the dominant index finger. Using experimental data and simulated abnormal spirals, we designed a decision support system that accurately identifies potentially abnormal spirals.
Conclusions: We built and validated a robust digital implementation of the Archimedes spiral exam and recommend a sensitive and specific workflow on the basis of data gathered from healthy volunteers.
Recommended Citation
Magee, Rogan; Yang, Benjamin; and Ratliff, Jeff, "Trsper: a web-based application for Archimedes spiral analysis" (2021). Department of Neurology Faculty Papers. Paper 263.
https://jdc.jefferson.edu/neurologyfp/263
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Language
English
Comments
This article is the authors’ final published version in mHealth. The published version is available at https://doi.org/10.21037/mhealth-21-16. Copyright © Magee et al.
Publication made possible in part by support from the Jefferson Open Access Fund