A03-KB109 Artificial Intelligence Based Radiogenomic Diagnosis of Gliomas

From Multidisciplinary Computional Anatomy
Jump to: navigation, search


  • Primary Investigator
    Manabu Kinoshita (Osaka University, Division of Neurosurgery, lecturer)
  • Corporative Researcher
    Takufumi Yanagisawa (Osaka University Graduate School of Medicine, Associate Professor)


Gliomas are known to be rare cancer that occur in the central nervous system (CNS). Prognosis of this disease is still not satisfactory in spite of multimodal treatment including surgery, radiation and chemotherapy. It has become clear that genetic alterations within the tumor have great impact on the treatment outcome of this disease and molecular mechanism(s) involved in the development of this disease is tightly regulated by several truncal mutations. As a result, it is now mandatory to acquire genetic information of the tumor in order to achieve optimum treatment strategy. Currently, direct tissue sampling is the only method available to satisfy this requirement and a less invasive method should be developed in the future. The aim of this research is to expand radiogenomic analysis and develop a robust non-invasive, radiological study based molecular diagnostic technology for gliomas. As artificial intelligence including deep learning and convolutional neural network is now widely available, the present research will pursue integrating this novel technology. This research is expected to extend further more from brain tumors to other types of cancers, which could be a milestone for integrating radiological and genetic information.

Project Design