Skip to content Search
Search our website:

Medical Natural Language Processing

Medical Natural Language Processing

Supervisors: Dr David Weston and Prof Mark Levene

Progress in Natural language processing (NLP) has been developing rapidly over the past decade, mainly due to statistical and deep learning methods. In particular, pre-trained language models that have been trained on very large text corpora, are allowing researchers access to a vast treasure of knowledge, that would not be readily available otherwise. Moreover, researchers can fine-tune these models to any particular domain such as medical and, additionally, distil its knowledge into smaller more manageable networks.

Learning structured medical information from free text, which is available from many sources, is still a challenging task, and can, through machine learning, allow for the discovery of medical knowledge. Such mined knowledge can also enable the training of automated diagnostic models, which can assist practitioners and ultimately patients.

The PhD will push forward the state-of-the-art in this area, looking at novel and scalable solutions that will build on existing pre-trained language models. This will involve modelling the problem, designing a solution, and also implementing a proof of concept using available data sets.

Candidate Requirements:

We are looking for candidates with a strong background in data science/machine learning, who also have some experience in Python as an implementation tool.

Key Reference:

[1] A. Hasan, M. Levene, and D. Weston. Learning structured medical information from social media. Journal of Biomedical Informatics, Volume 110:103568, 2020.

Further details about the project may be obtained from:

Principal supervisor: Dr David Weston and Prof Mark Levene

Further information about PhDs at CSIS is available via this link.