Hamamoto Yoshihiko

本稿では、まず人工知能の根幹に係る一般パターン認識の問題を明らかにするとともに、その解決に有用な統計的パターン認識とそれを発展させた離散Bayes識別則を概説する。次に離散Bayes識別則の個別化医療への応用としての診断・予測問題と創薬における治験問題に言及する。

Application of cell therapy using stem cell into intractable diseases has been noted as the medicine in the next generation taking the place of organ transplantation. Bone marrow cells have been focused as a candidate of the source of this therapy. Liver failure is a terminal state of all advanced liver diseases and thousands of patients are dying of decompensated liver cirrhosis in every year in Japan. Liver transplantation must be the strongest weapon to relieve those patients. However, particularly in Japan, liver transplantation has not been popularized. In this sense, we thought of autologous bone marrow cells transplantation for decompensated liver cirrhosis. Basic studies using animal model revealed that transplanted bone marrow cells into syngenic mouse suffering from liver cirrhosis migrated into the damaged liver and transdifferentiated to the hepatocytes. Based upon the animal studies, clinical trial of this new cell therapy supported by the Ministry of Health, Labor and Welfare is ongoing. It is early to estimate this therapy. However, free from liver failure has been observed in some cases.