Adjunct Professor of Biostatistics
Projects for Translational Statistics
Recently, advanced medical therapies/drugs are based on advanced basic science research and efficacy can be evaluated using small/ limited patients especially for the case of rare diseases. It is important to efficiently detect efficacy however deriving statistically significant p-values through standard statistics is difficult. The program for translational statistics is aimed to research new statistical concept bridging basic sciences, medical practice and statistical methodology.
There are many rare disease and regenerative data bases, however most of the data have not been effectively used due to various problems such as large variability due to non-homogeneous patients, limited number of data and data collected using an inadequate primary endpoint. Using the idea of translational statistics and working with medical expertise, variability of data may be reduced by predicting which patients will respond to which therapy. Finding and incorporating the most reasonable endpoint is another key factor in reducing variability and increasing precision of the data quality.
Through collaboration with US and Japanese research organizations, it may be possible to access large epidemiological data such as the Tohoku Medical Megabank Organization (ToMMo) data. With access to large epidemiological data, the hypotheses generated with small data set using translational statistics may be confirmed.
Sc.D.,1991, Harvard School of Public Health