Advancing explainable human in the loop NLP analytics for clinical applications

Lead Supervisor
Dr Iain Marshall
Clinical Academic Fellow
School of Population Health and Environmental Sciences, King’s College London

Dr Angus Roberts (Senior Lecturer in Biostatistics, King’s College London)
Dr Petr Slovak (Lecturer in Computer Science, King’s College London)
Serge Umansky, PhD (Metadvice)

Industrial Partner

Project Details

This studentship is a collaboration between KCL and Metadvice Ltd, an early stage digital health company focused on building high quality AI-driven clinical decision support tools, and will focus on the applications and improvements of the latest natural language processing (NLP) tools and systems for the purpose of providing knowledge parsing interfaces for neural network-driven clinical knowledge networks and decision support systems. 

The project will be focused on extracting entities, and representing medical concepts and relationships between the entities, from free-format medical text. Subsequently, identified elements of text will be further referenced to the concepts from clinical terminology classifications such as SNOMED Clinical Terms collection and/or others. On this basis the algorithms will be further developed allowing to reprocess unstructured medical texts, such as doctors notes, other documents and potentially voice recordings, into a format that is easy to analyze and systematize, automatically creating the possibility to extract useful knowledge from the currently underused and stranded unstructured data in clinical electronic medical record (EMR) system database, and other similar sources. 

Outcomes of this work will be used to integrate and improve on solutions developed for the purpose of providing clinicians with AI-augmented clinical decision support software. Potential candidates should have a good understanding of IT systems, practical experience with machine learning, experience with Python, knowledge of NLP basics, and an interest in extracting medical knowledge for the purpose of helping improve clinical decision making.