By Ian Bennett
A new pilot and exploratory study program at the University of Florida Health Cancer Center awarded funding to two teams of UF researchers to apply artificial intelligence (AI) to cancer research. These teams will address serious cancer-related research problems with AI to understand more about hard-to-study areas and develop new technologies like novel therapeutics.
The UF Health Cancer Center pilot program receives crucial support from the state of Florida through the Casey DeSantis Cancer Research Act (Fla. Stat. § 381.915).
Rui Yin, Ph.D., an assistant professor in the department of health outcomes and biomedical informatics in the University of Florida College of Medicine, will lead a project to further study contrastive convolutional neural networks to predict high-confidence mRNA-miRNA pairs in colorectal cancer patients. Xiangyang Lou, Ph.D., a research professor in the department of biostatistics in the University of Florida College of Public Health and Health Professions, will lead a team using the AI capabilities at the University of Florida to improve the quantification accuracy of magnetic particle imaging.
“These two awards exemplify the Cancer AI Working Group‘s overarching mission of developing novel AI approaches and applications to tackle previously insurmountable cancer research challenges, bridging interdisciplinary expertise and supporting early-career investigators,” said Lu and Prosperi, the Cancer AI Working Group co-chairs.
Gene expression is a complicated process controlled by a group of small RNA molecules named miRNAs by binding with messenger RNA (mRNA) in the cell cytoplasm. Studies show that miRNA levels in cancer cells are usually different from the miRNA levels in healthy cells, making them a promising target for cancer treatment. The Yin lab recently identified miRNAs that could be targeted if they are paired with their target mRNAs. Through their new project, the team aims to develop a machine-learning framework to efficiently predict high-confidence miRNA-mRNA interaction pairs of TDMD in colorectal cancer patients to improve current therapeutic interventions. This AI tool could be applied to various types of cancer and accelerate the development of cancer therapeutics.
Magnetic particle imaging (MPI) is a novel way to directly image magnetic nanoparticles in tissues. Only discovered in 2001, it could be a valuable tool in cancer research because of its high sensitivity, high spatial resolution and high imaging speed. However, an image reconstruction algorithm must be used to convert the signal from the machine into an image. Substantial distortion can occur during the calculation of the mass of nanoparticles due to partial volume effect and other factors. The collaboration led by Lou will develop a new AI-based analytical tool for producing more accurate quantifications of signals received in the imaging process.
“The funding of two new projects represents the approach of the UF Health Cancer Center to bring all of the talents of the university to bear on cancer,” said Jonathan Licht, M.D., director of the Cancer Center. “Using advanced computational techniques available at UF, these projects can uncover new understanding in gene regulation networks and how they are upset in cancer and develop new imaging techniques to aid in cancer diagnosis.”
By using the AI capabilities at UF, the teams hope to gain valuable insights and discover more effective treatments to make significant long-term improvements in patient outcomes. The UF Health Cancer Center is proud to support projects like these in an effort to fulfill its mission to prevent, detect, treat and ultimately cure cancer while addressing the unique challenges of the cancer burden faced by the population it serves.
