Machine learning aids eczema diagnosis

16 December 2020 | Muriel Cozier

…researchers have identified sets of two or three genes that together could distinguish irritant from allergic skin reactions.

Using algorithms, researchers at Karolinska Institutet in Sweden, have identified markers that can differentiate between irritant eczema and contact allergy. Their findings have been published in the journal Proceedings of the National  Academy of Sciences

The researchers say that about 20% of the population of high-income countries are affected by contact eczema, a disease often associated with exposure to chemicals in the environment. The two types of contact eczema each have their own cause. Allergic contact eczema is caused by allergic reaction and non-allergic infant eczema, caused by chemical agents or physical factors. Each type requires a different treatment.

Making the correct diagnosis can be difficult for dermatologists, as the diseases present similar symptoms and the results of a patch test can be difficult to interpret.

However, using a machine learning technique linked to a tailored genetic algorithm, researchers have identified sets of two or three genes that together could distinguish irritant from allergic skin reactions. The results were replicable in an independent group of patients and in external datasets.

Nanna Fyhrquist corresponding author, researcher and group leader of the Institute of Environmental Medicine, Karolinska Institutet commented; ‘Our results show that there is considerable potential for the development of new diagnostic methods based on these biomarkers. The next step in the project entails a more exhaustive clinical validation of the markers and technical optimisation of the method in order to attain sufficient cost-effectiveness and speed to clinical purposes.’

DOI:10.1073/pnas.2009192117

Related links:

Show me news from
All themes
from
All categories
by
All years
search by