Medtex: Unlocking the value of medical narratives
An Artificial Intelligence (AI) technology platform to ‘unlock’ key information in medical narratives for clinical decision support and secondary use.
Medicine is an information intensive field. Doctors, specialists and allied health care professionals, as well pathology and diagnosis centres, all produce patient health records.
Patient records contain valuable information for clinical decision support and secondary use (e.g. population health monitoring and reporting). However, searching and analysing health data poses unique challenges.
Most patient health reports – including clinical examination reports, nursing notes, discharge summaries and death certificates – are recorded in unstructured fragmented free-text. The text tends to be abbreviated, grammatically incorrect or contain spelling errors.
Medical concepts are often expressed using differing vocabulary. For example, hypertension and high blood pressure are two terms for the same condition. Sometimes concepts are implicit: haemodialysis implies renal failure. This means that manual inspections and experience-based judgements from clinical staff are required to evaluate information in the patient’s clinical records.
Patient records need to be easily searchable, accessible and understandable.
Medtex is a semantic medical text analysis software based on Natural Language Processing (NLP) and Machine Learning. It turns medical narratives into structured data that can be easily stored, queried or rendered by most systems for use in health applications.
The software has been developed in partnership with healthcare practitioners from cancer registries and hospital radiology and emergency medicine departments.
Medtex standardises free text by identifying medical concepts, abbreviations, acronyms, shorthand terms, dimensions and relevant legacy codes. It relates key medical concepts, terms and codes using contextual information and report substructure. The software also uses formal semantics to reason with the clinical concepts, inferring complex clinical notions relevant to a health application.
Medtex has delivered effective automated health monitoring and reporting solutions, including:
- the analysis of pathology reports and death certificates to assess the incidence of cancer and the associated mortality rates in a timely manner,
- the analysis of pathology and radiology reports to support the reconciliation of test result findings with emergency department discharge records,
- the analysis of medical reports to provide capability for medical record searching and analytics,
- and, the analysis of medical forums to identify adverse drug reactions.
Medtex provides a simple way of easily and consistently using free-text clinical data to improve health outcomes for patients, ease the workload of clinical staff, boost the efficiency of the health system and provide a rich data set for research.
The Australian e-Health Research Centre (AEHRC) is CSIRO's digital health research program and a joint venture between CSIRO and the Queensland Government. The AEHRC works with state and federal health agencies, clinical research groups and health businesses around Australia.