AIと脳核医学：認知症診断の新たな展望

AI and nuclear neuroscience：new perspectives for dementia diagnosis

近年，人工知能（artificial intelligence：AI）技術の進歩は，医療を含むあらゆる分野に革新をもたらしています。核医学領域は，画像そのものが生体の機能を反映するという特性を有し，さらに画像のピクセル数も比較的少ないことから，AIによる解析との親和性が極めて高い分野といえます。

In recent years, advances in artificial intelligence（AI）technology have brought about innovations in all fields, including medicine. Nuclear medicine is a field with a high affinity for AI analysis, due to the fact that images themselves reflect biological functions and have a relatively small number of pixels. AI analysis methods can be broadly divided into “supervised learning” and “unsupervised learning”. The former involves learning using predefined labels, including techniques such as decision trees and support vector machines（SVMs）. Among these, deep learning is particularly promising. On the other hand, unsupervised learning aims to automatically explore data structures without using labels and is widely applied to brain image analysis. In this article, I will introduce my research on AI analysis, primarily in nuclear medicine, and outline the current status and future prospects for medical applications of AI technology.