Human Phenotype Ontology Mapper – Kadhim et al., 2026

University of Southampton

Powerful Research

Researchers developed an artificial intelligence tool called HPO Mapper, which translates information from clinical records into a standardised medical language that computers can analyse.

The team tested the tool using more than 80,000 coded diagnoses, imaging reports and laboratory reports from 1,837 people with inflammatory bowel disease. HPO Mapper accurately translated different forms of clinical information and converted 62% of previously unusable free-text entries into standardised terms. The research was supported by an SDE Driver Grant. This preprint was published in May 2026 and has not yet been peer reviewed.

Delivering Public Benefit

HPO Mapper could help researchers identify rare genetic causes of illness, support more accurate diagnosis and inform more personalised treatments. It does this by turning information from clinical records, including previously difficult-to-analyse free text, into standardised data that can be compared with genes associated with disease. Although tested using inflammatory bowel disease records, the tool could potentially be used across a much wider range of conditions.

Enabled by the Wessex SDE

NHS England funding provided through the Secure Data Environment programme enabled the team to develop and test HPO Mapper using NHS clinical data. This work has created a reusable tool that can turn complex clinical information into standardised data for future research, reducing the need for each study to solve the same problem from scratch. Although initially tested using inflammatory bowel disease records, HPO Mapper could support research across many different diseases, extending the value of the original investment.

Powerful Research

Delivering Public Benefit

Enabled by the Wessex SDE

Further information