Japanese research team expects to use new cancer diagnosis method
[의학신문·일간보사=정우용 기자] A technology that can greatly increase the identification accuracy of cancer stem cells using artificial intelligence (AI) has been developed.
A research team at the Department of Applied Biology at Tokyo Institute of Technology in Japan announced that they had published a research paper in the Internet version of the International Journal of Molecular Sciences, expecting that it could be used for new cancer diagnosis methods.
Cancer stem cells are known to be present in small amounts in cancer tissues and have an important relationship with the generation and promotion of cancer. Although the detailed situation has not yet been revealed, it has been found that it is reversible, has unique properties in metabolism and immunity, and affects resistance to cancer treatment.
Therefore, the development of cancer treatment methods targeting cancer stem cells and surrounding tissues responsible for their maintenance is expected. Cancer stem cells have characteristic expressed genes and proteins, and the detection method is an important technology for characterizing cancer stem cells.
The research team developed a technology to identify cancer stem cells and non-cancer stem cells using generative AI, and began developing technology to improve its accuracy.
As a means to determine whether each cancer stem cell maintains the properties of a stem cell, a cell image showing the expression of the marker gene ‘Nanog’ was acquired, and as a result of learning the generative system AI along with the cell image, the prediction accuracy was very poor. In the case of images, it was confirmed that the average accuracy was improved by about 5.7 times compared to the existing method.