Structured report data from medical text report

Abstract

Structured report data from a medical text report is disclosed. A method and apparatus for generating structured report data from a medical text report is disclosed. The method comprises acquiring a medical text report comprising one or more natural language statements; analysing, using a computer implemented text analysis process, the medical text report to determine, for each natural language statement, one or more labels for the natural language statement; and generating structured report data comprising the determined one or more labels each in association with natural language data from the natural language statement to which the label corresponds. The computer implemented text analysis process comprises, for each natural language statement: determining, for each of the one or more words of the natural language statement, and using word embeddings, a vector representing the word; and determining, based on the determined one or more vectors, and using a text classification model, the one or more labels associated with the natural language statement.

Description

technical field [0001] The present invention relates to structured report data, and more particularly to generating structured report data from medical text reports, such as radiology reports. Background technique [0002] Medical text reports, such as radiology reports, are used in the medical field to communicate information between medical professionals. For example, a patient's physician may refer the patient for certain medical imaging, such as computed tomography angiography (CTA). Radiographic technologists then perform medical imaging on the patient, and radiologists then read or interpret the resulting images to produce text-based radiology reports of their findings. A medical discovery may be defined, for example, as an observation of a patient by a physician, eg, as a result of an examination, imaging, or other test of the patient. The radiology report is then provided to the referring physician. The referring physician can then make a diagnosis of the patient'...

AI Technical Summary

Problems solved by technology

The free-form and / or heterogeneous nature of reporting can make updating, integrating and or sharing information across databases difficult and / or inefficient

Furthermore, this may make performing further analysis (i.e., computer-implemented analysis and / or searching of information contained within medical text reports) difficult and / or inefficient (e.g., computationally expensive)

For example, based on free-form medical text reports, performing quantitative analyzes of specific patient groups or individual patients, or for population health analysis, or cross-sectional analysis to identify high-risk patients and / or improve population health insights, may is difficult and / or inefficient

Images