Professor Shangming Zhou’s research focuses on artificial intelligence (AI) and statistics in health and biomedical informatics: data-driven health-related studies using techniques, such as machine learning/deep learning, natural language processing, computational intelligence (artificial neural networks, fuzzy logic, nature-inspired computing etc), statistical analytics, and data mining.
Two current projects are mining data from electronic patient codes. First, deep phenotyping is being used to predict the development of colorectal cancer stages, including the potential return of cancer after treatment and associated multimorbidity. By revealing connections and interactions between phenotypic factors, the researchers seek ultimately to explore how and why cancer affects people differently and suggest how treatment and prevention could be individualised for sufferers. Second, AI is being applied to electronic records to build evidence for the safe use of medications to assist practitioners in improving their medication-use systems to prevent medication errors and patient harm.
In another project and in collaboration with the charity SUDEP Action and SUVO company, Professor Zhou and his team are using patient self-generated data via the Epilepsy Self-Monitoring (EpSMon) app to identify the risks and health outcomes of childbearing women. Through this research, they hope to gain an understanding of how digital technologies might be used to improve the wellbeing of expectant mothers who suffer from seizures.
Supported by Health Data Research UK and an international partner, Professor Zhou and his team are also developing data-driven solutions to identify the complex interactions between the socioeconomic, cultural and environmental factors that contribute to individual- and population-level health outcomes. In particular, the team advances local modelling technology to explore interactions of these factors at a micro-level across different sub-regions of data space so that they can effectively identify those sub-populations. They hope this will provide important insights into targeted policy development and intervention.