Rey Figure Project

Within the project, ”Towards Machine-based scoring of Neuropsychological Screening Tests“, we aim to create new digital methodologies and technologies to objectify the diagnostic assessment of cognitive functions. Besides potentially providing valuable digital tools for clinicians, the project will generate a unique data source that may be used in future studies in the field of mental health and in the field of computer science. The project is marked by a strong interdisciplinary collaboration between our lab, the Computer Science Department of the ETH Zurich (Prof. Ce Zhang), Rico Leuthold (https://smartcode.ch/wo/) and several international psychiatric clinics (Prof. Peter Brugger, University Hospital Zurich, Switzerland; Prof. Oskar Jenni, Children’s Hospital Zurich, Switzerland; Dr. Juan Arango, BioCruces Health Research Institute. Cruces University Hospital, Spain; Dr. Federica Scarpina, University of Torino, Italy; Prof. Qianhua Zhao, Huashan Hospital, China, Dr. Tino Zähle, University of Magdeburg). With this project, we have been awarded with very prestigious funding (SNF Bridge Discovery). We will take next steps to initiate a spin-off with the help of Unitectra (Innovation Hub Zurich).

Neurological and psychiatric disorders are among the most common and debilitating illnesses across the lifespan. Currently, neuropsychologists typically use paper-pencil tests to assess individual neuropsychological functions and brain dysfunctions, including memory, attention, reasoning and problem-solving. Most neuropsychologists around the world use the Rey-Osterrieth complex figure (ROCF) in their daily clinical routine: this provides insights into a person’s nonverbal visual memory capacity, which is key to quality of life and typically changes in aging. To obtain a score for a person’s visual memory capacity, a trained clinician inspects the reproduced ROCF drawing and tracks deviations from the original figure to form a final score. Specifically, the figure is split into 18 identifiable areas, each of which is considered separately and marked on the accuracy of its position and the distortion exhibited. Currently, the quantitative scoring is undertaken manually in a subjective manner and has been criticized for its unreliability in a number of publications. The scoring might vary as a function of motivation and tiredness or because the clinicians may be unwittingly influenced by interaction with the patient. Moreover, inexperienced clinicians are prone to potential erroneous scoring.

An objective and reliable machine-based scoring system is in great demand and would remove a time-consuming and tedious task from skilled clinicians (some schemes can take up to 15 minutes per figure). We have developed a machine-based automated quantitative scoring system, which outperforms both amateur raters and clinicians. If you have drawings from previous studies with a data usage agreement or if you are interested in a collaborating with us, please contact us: nicolas.langer@uzh.ch