Approach to sparse modeling based on compressed sensing

2017 Fiscal Year Final Research Report

• Project Area: Initiative for High-Dimensional Data-Driven Science through Deepening of Sparse Modeling
• Project/Area Number: 25120008
• Research Category: Grant-in-Aid for Scientific Research on Innovative Areas (Research in a proposed research area)
• Allocation Type: Single-year Grants
• Review Section: Complex systems
• Research Institution: Kyoto University
• Principal Investigator: Tanaka Toshiyuki 京都大学, 情報学研究科, 教授 (10254153)
• Co-Investigator(Kenkyū-buntansha): 池田 思朗 統計数理研究所, 数理・推論研究系, 教授 (30336101) ; 大関 真之 東北大学, 情報科学研究科, 准教授 (80447549)
• Project Period (FY): 2013-06-28 – 2018-03-31
• Keywords: 圧縮センシング / スパースモデリング / ベイズ統計 / 核磁気共鳴

Outline of Final Research Achievements

In collaboration with the Medical Science team, Planetary Science team and others, we have developed image reconstruction methods from fewer numbers of measurement data in magnetic resonance imaging (MRI) and very long baseline interferometry (VLBI), in which compressed sensing has been utilized. As an example, using one of the proposed methods, we visualized spatio-temporal dynamics of metabolism of glucose injected into a tumor-bearing mouse in vivo via magnetic resonance spectroscopic imaging (MRSI), where one can observe spreading of the injected glucose over the whole body and conversion into lactate within the tumor tissue (the Warburg effect). Without compressed sensing, it takes several hours to acquire data even for a single frame, making measurement of spatio-temporal dynamics impossible. We have thus proved feasibility of the concept that with compressed sensing one can visualize spatio-temporal dynamics of biochemical processes in body non-invasively.

Free Research Field

情報理論