System and method for language agnostic patient discovery

A patient discovery system for analysing a database of clinical data containing clinical text associated with a plurality of patients, the patient discovery system comprising:

• a language adaptor circuit to receive a patient discovery configuration in a query language and identify the query language, wherein the patient discovery configuration is used to automatically determine one or more patients to be flagged for further clinical investigation;
• a data transformation circuit to receive the clinical text from the database and produce transformed clinical data for each of the plurality of patients;
• wherein the clinical text is in one or more source languages that are different to the query language of the patient discovery configuration;
• wherein the language adaptor circuit is agnostic to the one or more source languages and the query language, and receives the transformed clinical data, and detects the source language of the transformed clinical data, and uses the received transformed clinical data and the received patient discovery configuration s to produce a coupled representation;
• and
• a patient discovery circuit for receiving the coupled representation, and automatically determining from the coupled representation if the clinical text from the database of clinical data for a patient meets requirements of the patient discovery configuration in the query language, and if the requirements of the patient discovery configuration are met the patient is added to a flag list for further investigation.