New insight in spiral drawing analysis methods – Application to action tremor quantification

• Published in: Clinical neurophysiology Vol. 128; no. 10; pp. 1823 - 1834
• Main Authors: Legrand, André Pierre, Rivals, Isabelle, Richard, Aliénor, Apartis, Emmanuelle, Roze, Emmanuel, Vidailhet, Marie, Meunier, Sabine, Hainque, Elodie
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.10.2017; Elsevier
• Subjects: Adult; Aged; Computers, Handheld - standards; Data Analysis, Statistics and Probability; Diagnosis, Computer-Assisted - methods; Diagnosis, Computer-Assisted - standards; Empirical mode decomposition; Essential tremor; Essential Tremor - diagnosis; Essential Tremor - physiopathology; Female; Handwriting; Humans; Male; Middle Aged; Motor Skills - physiology; Physics; Reproducibility of Results; Spiral drawing analysis; Unravelling; Young Adult; Empirical mode decomposition; Spiral drawing analysis; Essential tremor; Unravelling; Spiral drawing analysis Essential tremor Unravelling Empirical mode decomposition
• DOI: 10.1016/j.clinph.2017.07.002
• ISSN: 1388-2457; 1872-8952

•New methods for computerized spiral analysis are proposed, which outperform the usual velocity method.•They provide new insight into tremor amplitude and frequency variations.•They were implemented on a computer via a digitized tablet. Spiral drawing is one of the standard tests used to assess tremor severity for the clinical evaluation of medical treatments. Tremor severity is estimated through visual rating of the drawings by movement disorders experts. Different approaches based on the mathematical signal analysis of the recorded spiral drawings were proposed to replace this rater dependent estimate. The objective of the present study is to propose new numerical methods and to evaluate them in terms of agreement with visual rating and reproducibility. Series of spiral drawings of patients with essential tremor were visually rated by a board of experts. In addition to the usual velocity analysis, three new numerical methods were tested and compared, namely static and dynamic unraveling, and empirical mode decomposition. The reproducibility of both visual and numerical ratings was estimated, and their agreement was evaluated. The statistical analysis demonstrated excellent agreement between visual and numerical ratings, and more reproducible results with numerical methods than with visual ratings. The velocity method and the new numerical methods are in good agreement. Among the latter, static and dynamic unravelling both display a smaller dispersion and are easier for automatic analysis. The reliable scores obtained through the proposed numerical methods allow considering that their implementation on a digitized tablet, be it connected with a computer or independent, provides an efficient automatic tool for tremor severity assessment.