Standardizing a Psychological Test

Nicolas Langer of the Department of Psychology at UZH and colleagues are developing a web application for the automated interpretation of the Rey-Osterrieth complex figure, a widely used neuropsychological test. The tool might save a single neuropsychological division up to 1500 hours or 180 workdays per year. We talked with him about this exciting project for which he received BRIDGE Discovery funding.

ZNZ News: What are the aims of your BRIDGE Discovery project?
Prof. Nicolas Langer: Most neuropsychologists around the world use the Rey-Osterrieth complex figure (ROCF) in their daily clinical routine: It provides insights into a person’s nonverbal visual memory capacity, which is key to independent living and an important measure of quality of life. A trained clinician inspects the reproduced ROCF drawing and tracks deviations from the original figure to form a score. The goal of our project is to develop a machine-based scoring system for the ROCF that has sufficient prediction accuracy and functionality allowing us to bring a corresponding web application on the market.

Why is machine-based scoring of the neuropsychological test needed? 
Currently, the quantitative scoring is performed manually, is highly subjective and has been criticized in several publications for its unreliability. Our machine-based scoring system produces an objective, more accurate, and more reliable result and takes a time-consuming and tedious task from a clinician. For each ROCF test, the scoring can take a clinician up to 15-45 minutes. We estimated that a single (bigger) neuropsychological division conducts more than 2000 complete ROCF tests per year. Clinicians are already excited about the time the application could save them and which they could then use for the interaction with their patients.

What are the biggest challenges in developing such a tool?
A prototype of the web-based tool is available, and we are working on improving accuracy and functionality. Our biggest challenge is to gather a large enough data set that spans the complete range of scores to quantify uncertainty. We aim for a total of 20’000 digitized ROCFs from various populations regarding age and diagnostic status (healthy or with psychiatric disorder). We use a crowdsourcing platform to get the required number of data points. Our algorithm already outperforms the scoring by clinicians, but for the real-life usability of a machine-based scoring system, the results need to be highly reliable.

Who is involved in this project?
Professor Ce Zhang from the Department of Computer Science at ETH and myself are the principal investigators. My post-doc Zofia Baranczuk is further involved. Additionally, we collaborate with multiple clinics internationally to get the necessary data from patients and to get feedback on the usability of the application. We work with Prof. Peter Brugger of the Rehabilitation Center Valens, Prof. Oskar Jenni of the Children’s Hospital Zurich, Noëmi Eggenberger & Olivia Zindel-Geisseler of the University Hospital Zurich, Dr. Juan Arango-Lasprilla of the BioCruces Health Research Institute in Spain, Dr. Federica Scarpina of the University of Torino in Italy, Prof. Qianhua Zhao of the Huashan Hospital in China and Dr. Kenneth Schuster of the Child Mind Institute in the USA.

What will be the impact of your work?
Although the application can be used with patients of all ages, the aging population poses a growing challenge to society as the prevalence of physical and cognitive decline will increase. It will become increasingly important to develop cost-effective digital technologies, which could serve as a prescreening for cognitive decline. Already now, in the Canton of Zurich alone, there are over 18 (neuro-)psychological units using the ROCF test on a daily basis. In addition, there are many private practices in Zurich that offer psychological assessments, the majority of which include the ROCF test. Because the ROCF test is used worldwide and is independent of language, the impact of our work can be large.
On a broader perspective, the gained expertise in the digitization of neuropsychological tests and their scoring using crowdsourcing may be adapted and applied to various other neuropsychological tests.

Further Reading

  • Prof. Nicolas Langer, Department of Psychology, Methods of Plasticity Research web page
  • Rey Figure Project web page

BRIDGE is a joint program conducted by the Swiss National Science Foundation (SNSF) and Innosuisse – the Swiss Innovation Agency. It offers new funding opportunities at the intersection of basic research and science-based innovation, thereby supplementing the funding activities of the two organizations. See website