2019/09/05
Keio University
Sanyo-Onoda City University
A group of researchers, including Katsutoshi Nishimoto (at the time), Associate Professor Akira Funahashi, Yuta Tokuoka, and Assistant Professor Takahiro Yamada of the Department of Biosciences and Informatics, Faculty of Science and Technology, Keio University, and Professor Noriko Hiroi of Sanyo-Onoda City University, has developed an AI that can predict the future direction of movement from current cell images by leveraging the property that cells change their shape before migrating. As a result, it was shown that the implemented AI can predict the future direction of movement in four directions (upper left, upper right, lower left, and lower right) for cells such as mouse fibroblasts with an accuracy of over 80%. Furthermore, by examining the features within the images that the AI learned, the researchers showed that the AI predicts the future direction of movement based on cell protrusions, which are known to form at the front during cell migration, and the characteristic shape of the cell's trailing edge. This revealed that the implemented AI makes future predictions based on biological evidence. These results indicate that AI is useful for predicting future states from current cell images, and this is expected to have applications in areas where future prediction is strongly required, such as cancer prognosis in the medical field.
Prior to publication in the academic journalPLOS ONE, an online early version of this research was published on Wednesday, September 4, 2019 (US time).
Please see below for the full press release.