Objectives and general architecture of the project
OBJECTIVE 1: Development of text-based technology to support diagnosis and prevention.
- Medical Information Extraction (Medical Entity Recognition, Temporal and Domain Relationships Detection)
- Prevention/diagnosis pattern discovery from semantic networks constructed from extracted information
OBJECTIVE 2: Development of text-based technology to support healthcare institution management.
- Classification of medical documents according to medical coding systems, both standardized and not (ICD-10 codes, procedures, ...)
- Comprehensive inventories of certain kinds of diseases
OBJECTIVE 3: Basic NLP and DM tools required to achieve the previous objectives.
- Development and adaptation of existing NLP tools to the medical domain, such as normalization tools (including acronym disambiguation) and anonymization tools
- Development and adaptation of data mining tools to identify relevant relations between medical concepts, and to extract temporal patterns in patient historical records
- Development of annotation resources (integration of different coding systems and ontologies in a unique platform with access via approximate search)
- Development of medical entity detection and classification techniques
- Development of medical entity negation and speculation detection techniques