development of bony lesion detection system for CT images and its clinical application

• Project/Area Number: 15K19775
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Radiation science
• Research Institution: The University of Tokyo
• Principal Investigator: Hanaoka Shouhei 東京大学, 医学部附属病院, 特任講師 (80631382)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Project Status: Completed (Fiscal Year 2017)
• Budget Amount: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000); Fiscal Year 2017: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000); Fiscal Year 2016: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000); Fiscal Year 2015: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
• Keywords: 医用画像工学 / 転移性骨腫瘍 / X線CT / 深層学習 / コンピュータ支援検出 / 医用画像処理 / セグメンテーション / 異常検知 / 骨転移 / コンピュータ支援診断 / 放射線診断学

Outline of Final Research Achievements

We developed a method to highlight bone metastases from CT datasets by using deep learning. When previous and current CT datasets are inputted, the algorithm can detect both osteolytic and osteoblastic metastases and output them. The algorithm is not a simple temporal subtraction, but it detects abnormalities using the estimated change (and deviation of the change) of the CT value both of which are calculated by a deep learning. Thus, the proposed method produces less false positives. Thanks to this, metastases are clearly shown in the proposed maximum intensity projection image, which helps radiologists to easily detect metastases.

Research Products

• [Journal Article] Landmark-guided diffeomorphic demons algorithm and its application to automatic segmentation of the whole spine and pelvis in CT images (2017)
  • Author(s): Hanaoka S, Masutani Y, Nemoto M, Nomura Y, Miki S, Yoshikawa T, Hayashi N, Ohtomo K, Shimizu A.
  • Journal Title: International Journal of Computer Assisted Radiology and Surgery Volume: 12(3) Issue: 3 Pages: 413-430
  • DOI: 10.1007/s11548-016-1507-z
  • Peer Reviewed / Acknowledgement Compliant
• [Journal Article] Automatic detection of vertebral number abnormalities in body CT images (2017)
  • Author(s): Hanaoka S, Nakano Y, Nemoto M, Nomura Y, Takenaga T, Miki S, Yoshikawa T, Hayashi N, Masutani Y, Shimizu A
  • Journal Title: International Journal of Computer Assisted Radiology and Surgery Volume: 印刷中 Issue: 5 Pages: 719-732
  • DOI: 10.1007/s11548-016-1516-y
  • Peer Reviewed / Acknowledgement Compliant
• [Presentation] 脊椎骨転移の検出における経時差分CTの有用性の検討 (2018)
  • Author(s): 星合 壮大
  • Organizer: 2018年度日本医学放射線学会 2018年4月13日 パシフィコ横浜
• [Presentation] Residual network-based unsupervised temporal image subtraction for highlighting bone metastases (2018)
  • Author(s): S. Hanaoka, T. Masumoto, S. Hoshiai, Y. Nomura, T. Takenaga, M. Murata, S. Miki, T. Yoshikawa, N. Hayashi, O. Abe
  • Organizer: CARS 2018 Computer Assisted Radiology and Surgery. June 20 - 23, 2018, Hotel NH Collection Friedrichstrasse, Berlin, Germany
  • Int'l Joint Research
• [Presentation] Fully automatic definition of anatomical landmarks in medical images: a feasibility study (2016)
  • Author(s): Hanaoka S, Nomura Y, Nemoto M, Miki S, Yoshikawa T, Hayashi N, Ohtomo K, Shimizu A
  • Organizer: Computer Assisted Radiology and Surgery 2016
  • Place of Presentation: Heidelberg, Germany
  • Year and Date: 2016-06-21
  • Int'l Joint Research
• [Presentation] Fully automatic definition of anatomical landmarks in medical images: a feasibility study (2016)
  • Author(s): Shouhei Hanaoka, Yukihiro Nomura, Mitsutaka Nemoto, et al.
  • Organizer: CARS (computer assisted radiology and surgery) 2016
  • Place of Presentation: Heidelberg, Germany
  • Year and Date: 2016-06-21
  • Int'l Joint Research