Non-structure to structured radiological report using Natural language processing.

Medical health records are essential tools for documenting patients' medical histories. Standardizing these records is crucial because it allows for efficient querying and retrieval of important patient information. Standardized records also facilitate the extraction of data about specific populations, which is valuable for designing effective public health policies. Imaging data, often stored in DICOM format, is a critical component of health records. Radiologists typically generate reports from this data. However, these radiological reports can vary significantly between institutions and often lack a standardized structure. Consequently, extracting meaningful medical information from these reports to inform health policies is extremely challenging. This project aims to develop a methodology for converting unstructured radiological reports into a structured format. This will enhance the usability of imaging data for policy-making and improve the overall effectiveness of health records