Filter All Projects

Project Details

Juette Astrid | Junior Fellow
2001-07-01 - 2001-09-30 | Research area: Cognition and Sociality
Automatic Measurement of Movement Quality in Depressive Illness

The quality of motion – how people move – changes under different conditions. In recent studies correlations of movement-quality with sex, personality-traits, mood, cognition, and psychological conditions could be demonstrated. Consequently people are also consistently judged on personality or attractiveness only by means of their movement.
Many different physiological factors contribute to the performance of movements. Some develop very early, e.g. anatomical constraints and brain-organization, others stay more flexible, like hormone-levels and neurotransmitter-concentrations. These factors can change during ontogeny but also in the adult individual, where they can either undergo short-time changes - e.g. reflecting moods - or long-time changes, reflecting psychological conditions or dirsorders.

In this work I will try to give an overview on the physiological influences on movement-quality, their ontogenetic development, and a comparison of "normal" and "abnormal" conditions using the example of depressive illness.
I have filmed 200 young women under controlled conditions, speaking about themselves, who then had to fill out two questionnaires, the "Beck´sche Depressions Inventar" and the "Eigenschaftswörterliste" to assess their depressive states. The first questionnaire is frequently used clinically to make first and following assessments on the degree of depression, the second one is a questionnaire on personality with a subscale for depressiveness as a personality trait. With the computer program "e-motion" the scenes have already been analyzed in respect to movement-quality (including e.g. speed, size, frequency, and complexity of the movements) of the whole body.
A traditional statistic analysis on correlations between the movement-data and the loadings on the depressiveness-scales will follow. In addition I will try to train an artificial neuronal network to automatically classify single subjects by their loadings on the questionnaire scales to create a model-free, easy-to-use method to e.g. monitor therapy-progress.

Depressiveness is considered to be a mainly neurotransmitter-mediated disorder with patients showing frequently described altered movement-patterns. However, these alterations are often considered to reflect the person´s mood and cognitive condition (ie changed movement-quality as a result of "sad mood", "no interest") rather than physiological parameters. The movement-patterns found in this study will therefore be discussed in respect to these two influences, physiology and cognition.