PhD research: Automatic detection of pulmonary tuberculosis in in chest X-ray images using trustworthy machine learning

Date
October 2024 to September 2028
Countries
Category
Keywords
covariate shift
chest X-ray
machine learning
tuberculosis
uncertainty quantification
Institutions
Jimma University (Ethiopia)
Research fields
Medicine and Health Sciences
Technology and Engineering

We plan to design a robust, reliable and trustworthy deep learning model to detect tuberculosis from chest X-ray images.
Our approach includes: i) implementing a benchmark model with context-aware features, ii) designing a framework for reliable uncertainty quantification of predictions, iii) investigating covariate shift in optimising training and testing datasets, and iv) conducting feasibility analysis for real-case deployment.