gms | German Medical Science

49. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds)
19. Jahrestagung der Schweizerischen Gesellschaft für Medizinische Informatik (SGMI)
Jahrestagung 2004 des Arbeitskreises Medizinische Informatik (ÖAKMI)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie
Schweizerische Gesellschaft für Medizinische Informatik (SGMI)

26. bis 30.09.2004, Innsbruck/Tirol

Assessment of Fine Motor Characteristics of Handwriting Movements Using a Multi Channel Digitizing Smart Pen

Meeting Abstract (gmds2004)

Suche in Medline nach

  • corresponding author presenting/speaker Christian Hook - University of Applied Sciences, Regensburg, Deutschland
  • Juergen Kempf - University of Applied Sciences, Regensburg, Deutschland
  • Georg Scharfenberg - University of Applied Sciences, Regensburg, Deutschland
  • Christian Lipfert - Sparrow Analytics GmbH, Regensburg, Deutschland

Kooperative Versorgung - Vernetzte Forschung - Ubiquitäre Information. 49. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 19. Jahrestagung der Schweizerischen Gesellschaft für Medizinische Informatik (SGMI) und Jahrestagung 2004 des Arbeitskreises Medizinische Informatik (ÖAKMI) der Österreichischen Computer Gesellschaft (OCG) und der Österreichischen Gesellschaft für Biomedizinische Technik (ÖGBMT). Innsbruck, 26.-30.09.2004. Düsseldorf, Köln: German Medical Science; 2004. Doc04gmds348

Die elektronische Version dieses Artikels ist vollständig und ist verfügbar unter: http://www.egms.de/de/meetings/gmds2004/04gmds348.shtml

Veröffentlicht: 14. September 2004

© 2004 Hook et al.
Dieser Artikel ist ein Open Access-Artikel und steht unter den Creative Commons Lizenzbedingungen (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.de). Er darf vervielf&aauml;ltigt, verbreitet und &oauml;ffentlich zug&aauml;nglich gemacht werden, vorausgesetzt dass Autor und Quelle genannt werden.


Gliederung

Text

Introduction

Fine motor function of the hand and fingers is an important indicator for disorders generated by illness or drug-treatment [1], [2], [3]. The analysis of handwriting movements is thus a common method for the diagnosis of psychomotor dysfunctions, the examination of disease progression, and the monitoring of patients during drug administration. In this study we present a novel pen-based device for the recording and evaluation of kinematic and dynamic properties of handwriting or drawing movements. Our BioMedPen measures pen tip coordinates x(t), y(t), tilt angles α(t), β(t) and pressures px(t), py(t), pz(t) applied along and vertical to the pen axis during writing on an arbitrary paper pad. In addition, grip pressure of the fingers holding the pen can be monitored. The pen is easy-to-handle, has a high resolution both in the time and frequency domain and transfers digitized time signals to an ordinary PC or notebook. Due to its ergonomic ballpoint pen design the BioMedPen is an ideal tool for applications in neuroscience, psychiatry and in pharmacological studies.

Methods

The assessment of multimodal kinematic and dynamic patterns of handwriting or drawing operations is the principal purpose and the main potential of the device presented in this paper. In this respect, our unique multifunctional BioMedPen system is superior to common digitizer tablets or pen based human computer input devices [4]. For a comprehensive monitoring of fine motor characteristics the pen is equipped with a diversity of sensors measuring

(i) the time dependent position x(t), y(t) of the tip on a paper pad by the technique of an optical computer mouse. The kinematics of handwriting movements can be calculated from these data.

(ii) the time dependent inclination angles α(t), β(t) of the handheld pen by an electrolytic tilt sensor. This sensing technique predominantly captures the fine neuromotor function of the fingers holding the pen.

(iii) the static and dynamic pressure in three dimensions, applied by the refill to vertically arranged strain gauges in x,y-direction and to a piezoelectric sensor in z-direction.

(iv) the pressure applied by the fingers gripping the pen, transferred by a fluid to a miniaturized pressure sensor. These data are suited for monitoring muscle spasm.

In this paper we will focus on a pen configuration [5] based on pressure and tilt sensors. A specific software for data acquisition and analysis developed in our mathematical department covers a broad range of computations, e.g. data preprocessing, feature extraction, classification, correlation and covariance, statistics and multivariate time-series analysis. Samples of handwritten items, text or drawing movements are stored online in a database, and patient-specific data are instantly evaluated. Global statistical features are extracted from Px(t), Py(t) and Pz(t) pressure signals and pen tilt angles α(t), β(t). For example, the duration, mean value, standard deviation, skewness, number of peaks, number of zeroes, sweep of polar angle, line integrals, radial loops, properties of the frequency spectrum, nonlinear approximation parameters, etc. are calculated. Derivatives of the time series are also analyzed. If required, regional and morphometric properties of the signal can be examined in more detail. Altogether, each sample is mapped to an n-dimensional feature vector, i.e. to a strongly compressed representation of the original handwritten item. Rapid algorithms are then used for various comparative measures, classification and association tasks (i.e. matching of test sample with prototypes or subsets) or for the monitoring of trajectories in n-dimensional space. Even biometrical traits can be assessed from the collected samples in order to classify or identify a particular human individual with respect to arbitrarily large reference databases of enrolled subjects [6].

Results

Preliminary field trials have been performed in our lab in order to validate the hardware and software components of the system, and to examine general motor function properties. Biometric traits were extracted from pressure signals recorded during handwriting of various patterns. The results demonstrate a high reproducibility of typical features computed from repeated samples collected under similar conditions. Measurements show an extraordinarily high sensitivity for differences and alterations in hand motor performance. Even minor deteriorations of fine motor behavior due to exogeneous factors (e.g. environmental conditions, alcohol) can be assessed. Comparison of data by means of a specific metrics (feature match) allows for the classification and association of individuals with respect to certain criteria, such as speed and fluency of handwriting, force amplitudes, velocity profiles, tremor frequency etc. Different levels of neuromotor activity can thus be monitored with the BioMedPen, based on a variety of sensor techniques.

Discussion

Our BioMedPen is a remarkably flexible system offering a wide range of application areas. Because of its user-friendly design and a high computation performance with immediate database access and online signal evaluation, BioMedPen is suited equally well for large clinical trials, the monitoring of drug effects in individual patients, and for various diagnostic and therapeutic purposes. The system is superior in many respects to current digitizing pen or tablet devices because it can be used with arbitrary paper pads, and it is equipped with a diversity of novel low cost sensors featuring a multimodaltool for neuromotor function analysis. We are convinced that our BioMedPen is an excellent platform for the development of standard methods in computer aided diagnostics, in pharmacology and in health care institutions. Preliminary experimental results are most promising, confirming our approach to develop a novel system for applications in biometrics, neurology and psychology. Particularly, the pen is an ideal tool for the classification and quantification of neuromotor disorders and fine motor deficiencies of patients under drug-treatment.


Literatur

1.
Mergl R, Tigges P, Schroter A, Moller HJ, Hegerl U. Digitized analysis of handwriting and drawing movements in healthy subjects: methods, results and perspectives. J Neurosci Methods 1999; 90(2): 157-69.
2.
Slavin MJ, Phillips JG, Bradshaw JL. Visual cues and the handwriting of older adults: a kinematic analysis. Psychol Aging 1996; 11(3): 521-6.
3.
Tucha O, Aschenbrenner S, Eichhammer P, Putzhammer A, Sartor H, Klein HE, Lange KW. The impact of tricyclic antidepressants and selective serotonin re-uptake inhibitors on handwriting movements of patients with depression. Psychopharmacology 2002; 159(2): 211-5.
4.
Hallgren, L. Use of digital pen technologies in home healthcare. University of Linköping, Thesis LIU-IMT-EX-328, Sweden 2002.
5.
Hook, C, Kempf, J, Scharfenberg, G. New Pen Device for Biometrical 3D Pressure Analysis of Handwritten Characters, Words and Signatures. Proceedings ACM Multimedia Berkeley, USA 2003: 38-44.
6.
Hook, C, Kempf, J, Scharfenberg, G. A Novel Digitizing Pen for the Analysis of Handwriting in Biometrics. In: Biometric Authentication Workshop. Prague 2004. Lecture Notes in Computer Science. Springer 2004.