Navigation and service

Use of cookies

By clicking on "Allow" you consent to the anonymous recording of your stay on the site. The evaluations do not contain any personal data and are used exclusively for the analysis, maintenance and improvement of our website. For further information on data privacy, please click on the following link: Data Privacy Policy

OK

Assessing measles elimination status through AI-assisted molecular surveillance

ZKI-PH_PhD2024_03 (ZKI-PH2 & ZKI-PH5 & FG12)

Background:

Measles is a considerable public health burden and comes with an enormous contagiousness and a case fatality rate of 1:1000. A global elimination program for measles (and rubella) virus has been called into action by WHO. The main feature for assessment of the elimination status on country basis is no longer the incidence or the vaccination coverage but the length of a transmission chain initiated by an imported case of measles. If certain variants are detected in a defined region no longer than 12 months in 3 subsequent years, measles virus (MV) can be considered as eliminated. MV is assigned to a transmission chain considering its genotype, i.e. a distinct variant of a MV genotype. Genotyping according to the WHO protocol relies on phylogenetic analyses of a highly divergent fragment of only 450 nt, which is part of the ORF encoding the nucleoprotein. Due to a recent reduction of the Measles diversity to only two circulating genotypes, assessment of elimination status has become more challenging.

Aim/s:

This project aims at using the genetic Measles virus data collected in Germany over the last more than 20 years to develop a method that robustly infers the Measles elimination status. This will require identification of imported cases using AI-assisted phylodynamic approaches, thereby enabling estimation of temporal transmission chain lengths. Visualization techniques will enable quick assessment of upcoming cases and outbreaks.

AI methods:

AI-assisted phylodynamics and visualization techniques will be used to distinguish imported cases from local mutations. Starting with Convolutional Neural Networks, the most suitable AI-method will be identified throughout the project. This project involves a variety of visualization methods. To visualize transmission chains, this will require methods that show relationships and connections between the data or show correlations between two or more variables.

Keywords:

Artificial Intelligence, Measles virus, Transmission dynamics, Visualization, Public Health

Date: 07.03.2024