Multimorbidity, Care Processes, and Recovery Trajectories in Hospitalised Patients: A Natural Language Processing Approach
Supervisors:
Prof Alasdair MacLullich, Usher Institute (University of Edinburgh)
Dr Arlene Casey, Centre for Medical Informatics, Usher Institute (University of Edinburgh
Dr Clare MacRae, Centre for Population Health Sciences, Usher Institute (University of Edinburgh)
Prof Miguel O. Bernabeu, Centre for Medical Informatics, Usher Institute (University of Edinburgh)
Summary:
This innovative PhD project will leverage novel natural language processing (NLP) techniques to enhance our understanding of outcomes in hospital inpatients with multiple conditions (multimorbidity, or MM).
MM is a key factor in poor inpatient outcomes. Yet MM research with inpatients has to date largely relied on structured healthcare data. This means that the role of obviously important clinical data such as acute symptoms, mobility barriers, and physiotherapy engagement, has not been studied, because such information is recorded only in free-text notes.
In this interdisciplinary project, supervised by clinical, computing science and healthcare analytics experts, you will have access to a rich, multimodal, dataset (DataLoch). You will develop new methods of extracting information from free-text notes using advanced NLP techniques. You will integrate this with structured data to identify new relationships between MM, inpatient care variables, and outcomes such as mortality and readmissions.
Working at the intersection of AI, public health, and clinical research, you will gain in-demand skills in NLP, data science, and health informatics. You will also engage directly with patients and clinicians, ensuring meaningful impact, ethical transparency, and clinical relevance.
This is a rare and exciting opportunity to be at the forefront of digital health innovation.