Researchers in Kyoto have unveiled a new method for detecting pancreatic cancer using microRNAs and artificial intelligence. Pancreatic cancer is often diagnosed in later stages due to the difficulty of early detection, leading to lower survival rates; however, this new approach shows potential for earlier diagnosis and treatment. The team, consisting of members from Kyoto University and Arkray Inc., worked on developing a tumor identification model through automated machine learning, which could revolutionize the early detection of pancreatic cancer.
Akihisa Fukuda, an associate professor from Kyoto University, highlighted the significance of this new method in identifying pancreatic cancer at an earlier stage with minimal damage to the body. The study focused on microRNAs, short fragments of RNA responsible for genetic regulation, selecting common types to combine with a tumor marker for creating a diagnostic model. Data collected from healthy individuals and pancreatic cancer patients was used to train an AI system for automated machine learning to refine the cancer detection model.
The newly developed method showed a significant increase in the detection rate, particularly in Stage 1 cases, compared to using the tumor marker alone. Positive readings for pancreatic cancer increased from zero to 50 percent in the very early stages of the disease where detection is usually challenging. The team's research outcomes were published in the British Journal of Cancer, demonstrating the effectiveness of their method in identifying less advanced pancreatic cancer. Arkray Inc. aims to introduce the prototype to the market for further research within one to two years, potentially offering a groundbreaking approach to early pancreatic cancer detection and treatment.