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

2020 Fiscal Year Research-status Report

Real-time adaptive control of robots for microsurgery: towards submillimeter accuracy without added sensors

Research Project

Project/Area Number 19K14935
Research InstitutionThe University of Tokyo

Principal Investigator

MarquesMarinho Murilo  東京大学, 大学院工学系研究科(工学部), 助教 (70837468)

Project Period (FY) 2019-04-01 – 2022-03-31
Keywords手術ロボット / 適応制御 / 画像処理 / ロボティクス / センサ融合
Outline of Annual Research Achievements

The online calibration of surgical instruments using endoscopic images is being investigated.
(1) The method for recognizing and tracking the shaft of surgical instruments on endoscopic images using deep learning was improved. The method was extended to rigid instruments. A preliminary evaluation using real data was performed.
(2) Several versions of the adaptive controller were derived, and a version with proper stability proof has been developed. Preliminary experiments show promise.
(3) A methodology was developed for the generation of realistic CG data based on photorealistic rendering. That development is being used to improve the deep learning algorithms.
(4) The robot programming library DQRobotics has been updated with improved control algorithms and a draft version of the adaptive controller.

Current Status of Research Progress
Current Status of Research Progress

1: Research has progressed more than it was originally planned.

Reason

From April 2020, the adaptive control algorithm has been studied in depth and experimental results have been obtained around September 2020. The algorithm has been further improved and state-of-the-art results have been obtained in March 2021.
Around May 2020, a framework for the generation of synthetic images using a virtual reality simulator was developed. The development of this framework was not initially planned and is a great addition to the project.
During fiscal year 2020, the instrument tracking algorithm has been improved and extended to the estimation of the pose of rigid instruments. It has also been preliminarily validated using real data our February-March 2021.
The DQ Robotics library has been updated, improved, and a draft of the adaptive controller code has been added during fiscal year 2020.

Strategy for Future Research Activity

The research will continue as originally planned.
- Further improve the instrument tracking algorithm for better accuracy.
- Integrate the instrument tracking algorithm with the adaptive controller.
- Validate the proposed algorithm in experiments using the SmartArm surgical system.

Causes of Carryover

With the limitation of transit given the ongoing pandemic, the participation in conferences which were originally planned were not feasible in 2020 and might not be feasible in 2021.
Nonetheless, the research itself has progressed in ways not initially planned, and the research will benefit from better computational systems, cameras, and devices in general that can be purchased with this funding.
Moreover, the funding will be used to pay for the publication costs of the currently unpublished results.

  • Research Products

    (7 results)

All 2020 Other

All Journal Article (2 results) (of which Int'l Joint Research: 2 results,  Peer Reviewed: 2 results) Presentation (4 results) (of which Int'l Joint Research: 2 results) Remarks (1 results)

  • [Journal Article] DQ Robotics: A Library for Robot Modeling and Control2020

    • Author(s)
      Adorno Bruno Vilhena、Marques Marinho Murilo
    • Journal Title

      IEEE Robotics & Automation Magazine

      Volume: In press Pages: 0~0

    • DOI

      10.1109/MRA.2020.2997920

    • Peer Reviewed / Int'l Joint Research
  • [Journal Article] The effects of different levels of realism on the training of CNNs with only synthetic images for the semantic segmentation of robotic instruments in a head phantom2020

    • Author(s)
      Heredia Perez Saul Alexis、Marques Marinho Murilo、Harada Kanako、Mitsuishi Mamoru
    • Journal Title

      International Journal of Computer Assisted Radiology and Surgery

      Volume: 15 Pages: 1257~1265

    • DOI

      10.1007/s11548-020-02185-0

    • Peer Reviewed / Int'l Joint Research
  • [Presentation] Single Shot Pose Estimation of Surgical Robot Instruments' Shafts from Monocular Endoscopic Images2020

    • Author(s)
      Masakazu Yoshimura, Murilo M. Marinho, Kanako Harada, Mamoru Mitsuishi
    • Organizer
      2020 IEEE International Conference on Robotics and Automation (ICRA)
    • Int'l Joint Research
  • [Presentation] Towards the Pose Estimation of Surgical Instruments' Shafts from Synthetic Endonasal Images of a Head Phantom2020

    • Author(s)
      Shimon Takechi, Murilo M. Marinho, Saul A. H. Perez, Kanako Harada, Mamoru Mitsuishi
    • Organizer
      16th Asian Conference on Computer Aided Surgery (ACCAS)
    • Int'l Joint Research
  • [Presentation] Single-Shotベースの画像処理を用いた物体の3次元位置姿勢推定2020

    • Author(s)
      吉村仁和, マルケスマリニョ・ムリロ, 原田香奈子, 光石衛
    • Organizer
      第29回日本コンピュータ外科学会大会
  • [Presentation] Segmentation of Endonasal Robotic Instruments in a Head Phantom using Deep Learning and Virtual-Reality Simulation2020

    • Author(s)
      Saul Alexis Heredia Perez, Murilo Marques Marinho, Kanako Harada, Mamoru Mitsuishi
    • Organizer
      ロボティクス・メカトロニクス 講演会2020 (Robomec)
  • [Remarks] DQ Robotics Website

    • URL

      https://dqrobotics.github.io

URL: 

Published: 2021-12-27  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi