Use Case: Archive Digitisation

University Childrens Hospital: Opening up a Worldwide Unique Data Set With Intelligent Document Processing (IDP)

Overview

As part of the Zurich Longitudinal Studies, the University Children's Hospital Zurich studied the physical, motor, cognitive and social development of more than 1'000 children since the 1950s.

In this pilot project, the feasibility of digital data extraction from questionnaires was tested, and the expected effort was hypothesized.

KiSpi

Task

Pilot project for the digitisation and extraction of longitudinal studies.

Scope

5'000 questionnaires

Services

ICR (Intelligent Character Recognition) and data extraction.

„By working with Acodis, extraordinarily valuable insights can be processed and made available for future research. A manual analysis would take years. ”

Prof. Oskar Jenni, Head of Zurich Longitudinal Studies

Starting Point

Since the 1950s, the University Children's Hospital Zurich has been studying the physical, motor, cognitive and social development of more than 1,000 children as part of the Zurich Longitudinal Studies. The study participants were regularly examined from birth until young adulthood. The resulting data set of the Zurich Longitudinal Studies (ZLS) is unique worldwide and will therefore be processed and made accessible for future research.

The data on child development was documented by hand on countless different examination protocols and questionnaires, resulting in a data archive with more than 750,000 sheets of paper. Since 2019, the data has also been supplemented by a further examination point in adulthood. The ZLS will thus become a life span study: a unique opportunity to examine health and development over the entire life span.

Ready to extract data from your studies?

Our Approach

With the pilot project, the feasibility of digital data extraction from hand-filled questionnaires was tested and the expected effort was hypothesized. To enable the recognition of handwriting, Acodis created a training set based on 5,000 pre-digitized questionnaires. Ultimately, an ICR (Intelligent Character Recognition) system could be created specifically for the questionnaires of the ZLS. The specialized ICR will then be used to recognize the numbers on all questionnaires in the pilot.

In a later step, the structure of the respective questionnaire was automatically recognized, and the data, structured. The digitized content can then be post-processed in the customized editor, taking various factors into account. Validation rules, such as statistical plausibility, are applied to maximize efficiency in the review. Automatically adapted data schemes for the different types of questionnaires also enable a clear and structured listing of the evaluated data.

Customer Benefits

Processing Handwriting

Save Time & Resources

Enabling Manual Review & Correction

Outcome

With the digitisation and data extraction platform, the University Children Hospital Zurich can sustainably preserve this unique data from over half a century of research for the environment by enabling manual review and correction. The setup system enables them to also save time and resources.

The pilot project was successfully completed and the data set will be continuously expanded in the coming months until all relevant questionnaires have been readout.

Nothing Replaces a Conversation

We are looking forward to talking to you.