Developments of diagnostic AI algorithms for renal tumor images

• Project/Area Number: 18K15635
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 52040:Radiological sciences-related
• Research Institution: Okayama University
• Principal Investigator: Tanaka Takashi 岡山大学, 大学病院, 助教 (10745368)
• Project Period (FY): 2018-04-01 – 2021-03-31
• Project Status: Completed (Fiscal Year 2020)
• Budget Amount: ¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000); Fiscal Year 2020: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000); Fiscal Year 2019: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000); Fiscal Year 2018: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
• Keywords: Deep learning / 腎腫瘍 / CT / 画像診断 / 人工知能 / 深層学習 / 放射線医学

Outline of Final Research Achievements

This study evaluated the utility of a deep learning method with convolutional neural networks (CNNs) for determining whether a small solid renal mass was benign or malignant on multiphase contrast-enhanced CT. A deep learning method with CNNs allowed acceptable differentiation of small solid renal masses in dynamic CT images. However, a single deep learning model could not predict malignancy in all renal tumors of out study.
By preparing and adjusting the appropriate images and patients for training, we might be able to create more promising models for various specialized tasks.

Academic Significance and Societal Importance of the Research Achievements

本研究によりDeep learningによる腎腫瘍の診断法の一定の有用性が示され、Deep leaning技術を用いることで、画像評価者の経験や違いなどに影響のない、より均一な精度の検査を多くの患者に提供できる可能性が示された。また、同様の手法を応用すれば、腎臓以外の多くの腫瘤の診断への適応が拡がる可能性も示唆される。さらにこの解析法の普及および産学官連携により、国内でのより精度を高めたDeep learningソフトウェアの開発や、国内のビッグデータを用いたクラウドデータベース構築など発展的研究につながることが期待される。

Research Products

• [Journal Article] Imaging evaluation of hereditary renal tumors: a pictorial review (2021)
  • Author(s): Takashi Tanaka, Akira Kawashima, Yohei Marukawa, Takahiro Kitayama, Yoshihisa Masaoka, Katsuhide Kojima, Toshihiro Iguchi, Takao Hiraki, Susumu Kanazawa
  • Journal Title: Japanese Journal of Radiology Volume: -
  • NAID: 210000187555
  • Peer Reviewed
• [Journal Article] Differentiation of Small ( 4 cm) Renal Masses on Multiphase Contrast-Enhanced CT by Deep Learning. (2020)
  • Author(s): Tanaka T, Huang Y, Marukawa Y, Tsuboi Y, Masaoka Y, Kojima K, Iguchi T, Hiraki T, Gobara H, Yanai H, Nasu Y, Kanazawa S.
  • Journal Title: AJR Am J Roentgenol Volume: - Issue: 3 Pages: 1-8
  • DOI: 10.2214/ajr.19.22074
  • Peer Reviewed / Int'l Joint Research
• [Presentation] Getting Started with Artificial Intelligence: A Quick Start Guide for Radiologists (2020)
  • Author(s): Tanaka Takashi, Marukawa Yohei, Samura Kazuma, Kitayama Takahiro, Masaoka Yoshihisa, Kojima Katsuhide, Hiraki Takao, Kanazawa Susumu.
  • Organizer: Radiological Society of North America
  • Int'l Joint Research