データマイニング手法を用いた医薬品副作用発現情報のビジュアル化と解析，臨床応用

概要

論文の詳細を見る
Sources of drug information such as package inserts (PIs) and interview forms (IFs) and existing drug information databases provide primarily document-based and numerical information. For this reason, it is not easy to obtain a complete picture of the information concerning many drugs with similar effects or to understand differences among drugs. The visualization of drug information may help provide a large amount of information in a short period, relieve the burden on medical workers, facilitate a comprehensive understanding and comparison of drugs, and contribute to improvements in patients' QOL. At our department, we are developing an approach to convert information on side effects obtained from PIs of many drugs with similar effects into visual maps reflecting the data structure through competitive learning using the self-organizing map (SOM) technique of Kohonen, which is a powerful method for pattern recognition, to facilitate the grasping of all available information and differences among drugs, to anticipate the appearance of side effects; we are also evaluating the possibility of its clinical application. In this paper, this approach is described by taking the examples of antibiotics, antihypertensive drugs, and diabetes drugs.