Efficient & Effective Tools for Biomedical Relations Extraction
The aim of this final year project is to to develop a clinical decision support system (CDSS) that assists clinicians in diagnosis of rare or complicated diseases. The system is positioned to provide insights and updated researches as references for clinicians to make use of, but not replace a doctor’s specialty in diagnosis. Users can input patient’s information to the system, and receive relevant advice, such as paper suggestion, possible disease relation graphics, etc.
Rare and newly discovered syndromes are areas where even experienced doctors are uncertain about. Medical research papers are the main source of knowledge for the above-mentioned field. However, as the number of medical research papers accumulates each year, it has become more difficult and nearly impossible for doctors to follow every medical finding. The demand of this function of CDSS emerged.
In order to serve the purpose of decision support, the output of the system includes ranking of research paper suggestions, graphics that shows relations between patients’ clinical information and possible diagnosis or risks of diseases, etc. This allows clinicians to narrow down the wild search of possible diseases, and realize possibility of different diagnosis that was neglected due to lack of knowledge in new medical findings. By providing insights extracted from medical research paper, we hope to further align medical finding and clinical services more efficiently, with an objective to improve health care.