Development and prospective validation of a deep learning model to detect abnormal clinical laboratory measurements in the entire Finnish population

Description of the granted funding

One hundred forty-two million clinical laboratory tests were performed in Finland in 2022, making it the highest-volume healthcare service. Current approaches to blood testing in primary healthcare are opportunistic and do not consider the richness of health data or integrate genetic information. Importantly, we showed that individuals from a disadvantaged socio-economic background get tested less, while other individuals are over-tested. We hypothesize that AI can be successfully employed to identify individuals who are likely to have abnormal values. We propose to develop and prospectively validate an AI approach to identify individuals who would benefit most from laboratory testing. We will do that using the extensive health and genetic data resources. Finally, we will recontact 2,000 individuals with predicted poor kidney function. This will allow us to understand the quality of our approach, while potentially identifying individuals with underdiagnosed chronic kidney disease.
Show more

Starting year

2024

End year

2028

Granted funding

Andrea Ganna Orcid -palvelun logo
599 886 €

Funder

Research Council of Finland

Funding instrument

Academy projects

Decision maker

Scientific Council for Biosciences, Health and the Environment
12.06.2024

Other information

Funding decision number

361890

Fields of science

Biomedicine

Research fields

Biolääketieteet

Identified topics

public health, occupational health