Cancer Translational Research Team (https://aip.riken.jp/labs/goalorient_tech/cancer_transl/) at RIKEN AIP
Speaker 1 (Approx. 45 mins): Ryuji Hamamoto
Title: Medical AI research for clinical applications
Abstract: In recent years, machine learning and deep learning technologies have been applied to medical applications, and several AI-equipped medical devices have already been approved by the US FDA. In this context, the Cancer Translational Research Team’s mission is to apply machine learning and deep learning techniques to medicine, with a particular focus on the clinical implementation. As the main framework, we are promoting two projects, medical image analysis and omics analysis. With regard to the medical image analysis, the endoscopy project and the ultrasound project are currently being actively pursued, and we are aiming for early social implementation of both projects. In addition, we are currently conducting research on omics analysis, using machine learning and deep learning technologies to analyze large scale data on lung cancer in order to elucidate the molecular mechanisms of tumorigenesis and apply it to cancer treatment and drug discovery. In this seminar, I plan to present our current progress and our future strategy.
Speaker 2 (Approx. 30 mins): Masaaki Komatsu
Title: Development of a diagnostic ultrasound system using AI technology
Abstract: Compared to other modalities, ultrasonography is superior in simplicity, non-invasiveness, and real-time performance, and has become widespread in a wide range of clinical medical disciplines. On the other hand, ultrasound images are highly dependent on the skill of the examiner, such as manual scanning, and are easily affected by the acoustic shadowing of bones and other obstructions, so there are specific problems in controlling the accuracy of the images. Therefore, it is expected that the use of AI based on ultrasound imaging will solve the shortage of manpower and human errors of medical workers, and improve the quality and uniformity of medical care. In order to achieve this goal, it is necessary to accumulate highly robust basic technologies including data structuring and algorithm development using high quality clinical data. In this seminar, I plan to present our results to date and our strategy for clinical applications.
Speaker 3 (Approx. 30 mins): Ken Asada
Title: Multi-omics analysis for elucidating the molecular mechanisms of lung cancer
Abstract: At present, the promotion of Precision Medicine has become an international trend. On the other hand, the current method of predicting the most appropriate treatment based on limited genetic mutation information from diagnosis using targeted-gene panels (TGPs) is recently causing problems because the number of patients who can benefit from Precision Medicine is small. In order to realize true Precision Medicine, it is necessary to build a system that benefits more patients through multimodal analysis of multiple layers of omics data, including whole genome analysis, epigenome analysis, and proteomics analysis. Therefore, we are currently using machine learning and deep learning technologies to construct a multi-modal analysis system for information with different hierarchies such as genome information and epigenome information. These results will contribute to the elucidation of the molecular mechanisms of cancer, drug discovery, and the promotion of Precision Medicine. In this seminar, I plan to present our achievements to date and future strategies.