الفهرس | Only 14 pages are availabe for public view |
Abstract Many medical record systems are used to record important information (such as demographic information, patient symptoms, diagnostic results, results of examinations and tests, medical history, etc.) to accurately describe patient’s health status and many more additional features. This, in turn, makes it easier for physicians to decide an accurate diagnosis for patient. However, many physicians are strongly opposed to using any electronic health record system due to the large amount of data and procedures that must be entered into it because their main interest is the quick diagnosis and treatment of inpatients. Therefore, The goal of this thesis is to present a novel approach. that makes it easier for physicians to enter data about the symptoms that patients are complaining about, and to facilitate assistance in reaching the correct diagnosis for patient complaint, by providing an agent that patient will use to describe his/her medical complaint in a written textual description. Then the agent uses the entered description to extract the disease symptoms using natural language processing (NLP) techniques in preparation for their inclusion in the patient’s medical record, saving physicians the time and effort of entering data that should have been entered into their medical records. After that, the agent proposes to the attending physician the closest diagnosis whose symptoms are similar to the patient’s complaint. In addition, the agent presents medical information guidelines on the proposed diagnoses such as the common diagnostic tests and procedures. Thus, the agent assists the physician in quickly reaching the preliminary diagnosis closest to the patient’s complaint. |