• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Computer-aided detection development based on semi-supervised learning with medical chart information

Research Project

Project/Area Number 20K11944
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 61030:Intelligent informatics-related
Research InstitutionKindai University

Principal Investigator

Mitsutaka Nemoto  近畿大学, 生物理工学部, 講師 (10451808)

Project Period (FY) 2020-04-01 – 2023-03-31
Project Status Completed (Fiscal Year 2022)
Budget Amount *help
¥3,770,000 (Direct Cost: ¥2,900,000、Indirect Cost: ¥870,000)
Fiscal Year 2022: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2021: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2020: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Keywordsコンピュータ支援診断AI / 弱教師学習 / 医用画像 / 3D U-Net / 適応的閾値処理 / コンピュータ検出支援 / 病変領域教師ラベル / 胸部CT像 / 肺結節 / U-Net / 異常検知 / FDG-PET/CT像 / 医用画像診断 / 半教師あり学習 / カルテ情報
Outline of Research at the Start

病変領域を画素単位で指摘した病変領域教師ラベルデータは,コンピュータ検出支援:CADe(Computer-aided detection)システム開発に不可欠だが,同教師ラベルの生成・大量収集は極めて困難である。教師ラベルデータを持たないデータを機械学習に用いる半教師あり学習は,教師ラベル推定時の誤りに対する懸念がある。本研究では,病変領域教師ラベルデータが少数しか得られない環境において,カルテ情報を用いて高精度に推定した教師ラベルを効果的に用いたCADeシステム開発を確立する。

Outline of Final Research Achievements

In this research, we mainly proposed and experimentally investigated a method for automatically assigning highly accurate teacher labels on medical images without lesion region teacher labels by fusing lesion information from medical record data into a supervised machine learning framework.
The proposed method estimates lesion area teacher labels from extremely rough and weak teacher data, consisting of lesion position coordinates and lesion length diameter, by using multiple 3D U-Nets. Validation experiments by CT image data, including lung nodules, showed that the estimated teacher labels with good Dice coefficients were obtained for all three types of nodules: solid nodules, GGO nodules, and Mixed-GGO nodules.
We also investigated pixel discrimination, which is essential for lesion area extraction. A multi-species lesion detection process on FDG-PET/CT images based on unsupervised pixel anomaly detection was proposed and published as a journal article.

Academic Significance and Societal Importance of the Research Achievements

本研究は、画像診断支援AIシステム開発において不可欠なデータである"病変領域教師データ"の生成コストを大幅に削減することに寄与する可能性を秘めた研究である。この病変領域教師データは、専門知識を持つ放射線診断医が画素単位で手入力することでしか作成することができない。よって、病変領域教師データの作成および収集における人的・時間的コストは極めて大きい。本研究は、このようなデータ作成を少数に抑えることができ、かつ病院に蓄積された診断データを教師ラベルの作成作業無しに後ろ向き研究利用するための方策となりうる。この技術を用いることで診断支援AIシステムの開発速度向上が見込めることから、社会的意義は高い。

Report

(4 results)
  • 2022 Annual Research Report   Final Research Report ( PDF )
  • 2021 Research-status Report
  • 2020 Research-status Report
  • Research Products

    (18 results)

All 2023 2022 2021 2020

All Journal Article (2 results) (of which Peer Reviewed: 2 results) Presentation (16 results) (of which Int'l Joint Research: 7 results)

  • [Journal Article] Automatic detection of primary and metastatic lesions on cervicothoracic region and whole-body bone using a uniform machine-learnable approach for [18F]-FDG-PET/CT image analysis2022

    • Author(s)
      Nemoto Mitsutaka、Tanaka Atsuko、Kaida Hayato、Kimura Yuichi、Nagaoka Takashi、Yamada Takahiro、Hanaoka Kohei、Kitajima Kazuhiro、Tsuchitani Tatsuya、Ishii Kazunari
    • Journal Title

      Physics in Medicine & Biology

      Volume: 67 Issue: 19 Pages: 195013-195013

    • DOI

      10.1088/1361-6560/ac9173

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed
  • [Journal Article] A review on AI in PET imaging2022

    • Author(s)
      Matsubara Keisuke、Ibaraki Masanobu、Nemoto Mitsutaka、Watabe Hiroshi、Kimura Yuichi
    • Journal Title

      Annals of Nuclear Medicine

      Volume: 36 Issue: 2 Pages: 133-143

    • DOI

      10.1007/s12149-021-01710-8

    • Related Report
      2021 Research-status Report
    • Peer Reviewed
  • [Presentation] 画像診断支援システムの弱教師学習に向けたCT 像マルチウィンドウ画像解析による肺結 節教師領域ラベル自動推定2023

    • Author(s)
      村中皓紀、根本充貴、木村裕一、永岡隆、細田和史、大谷和暉、 吉田昂平、吉川健啓
    • Organizer
      第62回日本生体医工学会大会
    • Related Report
      2022 Annual Research Report
  • [Presentation] Pix2Pix画像スタイル変換を用いた教師無し異常検知によるFDG-PET/CT像上肺病変強調2023

    • Author(s)
      大谷和暉、根本充貴、甲斐田勇人、瀬川新、中前有香子、村中皓紀、吉田昂平、北島一宏、石井一成
    • Organizer
      第62回日本生体医工学会大会
    • Related Report
      2022 Annual Research Report
  • [Presentation] Study for detecting pulmonary nodules on FDG-PET/CT images by deep image generation-based anomaly detection with training small dataset2022

    • Author(s)
      A Segawa, M Nemoto, H Kaida, Y Kimura, T Nagaoka, H Yamaguchi, Y Nakamae, T Yamada, K Hanaoka, K Kitajima, T Tsuchitani, K Ishii
    • Organizer
      World Federation of Nuclear Medicine and Biology (WFNMB) 2022
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Detection of Bone Metastasis on FDG-PET/CT Images using Multi-step Anomaly Voxel Detection and Local Patch analysis with Unsupervised Deep Features and Image Textures2022

    • Author(s)
      H. Yamaguchi, M. Nemoto, H. Kaida, Y. Kimura, T. Nagaoka, T. Yamada, K. Hanaoka, K. Kitajima, T. Tsuchitani, K. Ishii
    • Organizer
      World Federation of Nuclear Medicine and Biology (WFNMB) 2022
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 深層画像生成技術を用いたFDG-PET/CT像異常検知による病変強調2022

    • Author(s)
      瀬川新、根本充貴、甲斐田勇人、山口明乃、木村裕一、永岡隆、山田誉大、北島一宏、石井一成
    • Organizer
      第61回日本生体医工学会大会
    • Related Report
      2022 Annual Research Report
  • [Presentation] Detection of bone metastases on FDG-PET/CT images by using two-step anomaly detection.2021

    • Author(s)
      H Yamaguchi, M Nemoto, H Kaida, Y Kimura, T Nagaoka, T Yamada, K Hanaoka, K Kitajima, T Tsuchitani, K. Ishii
    • Organizer
      CARS 2021 - Computer Assisted Radiology and Surgery
    • Related Report
      2021 Research-status Report
    • Int'l Joint Research
  • [Presentation] FDG-PET/CT画像に対する2種類の異常検知を用いたがん骨転移病変の自動検出2021

    • Author(s)
      山口明乃, 根本充貴,甲斐田勇人,木村裕一,永岡隆,山田誉大,花岡宏平,北島一宏,槌谷達也,石井一成
    • Organizer
      第61回日本核医学会学術総会
    • Related Report
      2021 Research-status Report
  • [Presentation] 多段の画素異常検知によるFDG-PET/CT上のがん骨転移候補検出2021

    • Author(s)
      山口明乃, 根本充貴,甲斐田勇人,木村裕一,永岡隆,山田誉大,花岡宏平,北島一宏,槌谷達也,石井一成
    • Organizer
      第40回日本医用画像工学会大会
    • Related Report
      2021 Research-status Report
  • [Presentation] 2種類のAI異常検知カスケードを用いたFDG-PET/CT像上がん骨転移検出2021

    • Author(s)
      山口明乃,根本充貴,甲斐田勇人,木村裕一, 永岡隆,山田誉大,花岡宏平,北島一宏,槌谷達也,石井一成
    • Organizer
      第60回日本生体医工学会大会
    • Related Report
      2021 Research-status Report
  • [Presentation] Detection of cerebral aneurysms on MR angiography using generated features by unsupervised deep learning for multiple 2.5-dimensional images2021

    • Author(s)
      K. Ushifusa, M. Nemoto, Y. Kimura, T. Nagaoka, T. Yamada, N. Hayashi
    • Organizer
      International Forum on Medical Imaging in Asia
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] A generalized image feature generation based on unsupervised deep learning with small scale normal dataset2020

    • Author(s)
      K. Ushifusa, M. Nemoto, Y. Kimura, T. Nagaoka, T. Yamada, N. Hayashi, A. Tanaka
    • Organizer
      Computer Assisted Radiology and Surgery
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] Automatic detection of cervical and thoracic lesions on FDG-PET/CT by organ specific one-class SVMs2020

    • Author(s)
      A. Tanaka, M. Nemoto, H. Kaida, Y. Kimura, T. Nagaoka, T. Yamada, K. Ushifusa, K. Hanaoka, K. Kitajima, T. Tsuchitani, K. Ishii
    • Organizer
      Computer Assisted Radiology and Surgery
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] A pilot study for transferring deep convolutional neural network pre-trained by local anatomical structures to computer-aided detection.2020

    • Author(s)
      M Nemoto
    • Organizer
      Computer Assisted Radiology and Surgery
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] 少数症例を用いた教師無し深層学習による病変検出画像特徴量の汎用的生成 -胸部CTを用いた実験的検証-2020

    • Author(s)
      牛房和之, 根本充貴, 木村裕一, 永岡隆, 山田誉大, 田中敦子, 林直人
    • Organizer
      日本医用画像工学会大会
    • Related Report
      2020 Research-status Report
  • [Presentation] One-class SVMを用いた異常検知によるPET/CT上の骨転移病変自動検出2020

    • Author(s)
      田中敦子,根本充貴,甲斐田勇人,木村裕一,永岡隆,山田誉大,牛房和之,花岡宏平,北島一宏,槌谷達也,石井一成
    • Organizer
      日本核医学会学術総会
    • Related Report
      2020 Research-status Report
  • [Presentation] one-class SVMによる画素悪性度の集中性を用いたFDG-PET/CT上の病変自動検出2020

    • Author(s)
      田中敦子,根本充貴,甲斐田勇人,木村裕一,永岡隆,山田誉大,牛房和之,花岡宏平,北島一宏,槌谷達也,石井一成
    • Organizer
      日本医用画像工学会大会
    • Related Report
      2020 Research-status Report

URL: 

Published: 2020-04-28   Modified: 2024-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi