📄 Towards a Simple Framework of Skill Transfer Learning for Robotic Ultrasound-guidance Procedures

Tsz Yan Leung 2 and Miguel Xochicale 1
1 University College London, and 2 King’s College London

Abstract

In this paper, we present a simple framework of skill transfer learning for robotic ultrasound-guidance procedures. We briefly review challenges in skill transfer learning for robotic ultrasound-guidance procedures. We then identify the need of appropriate sampling techniques, computationally efficient neural networks models that lead to the proposal of a simple framework of skill transfer learning for real-time applications in robotic ultrasound-guidance procedures. We present pilot experiments from two participants (one experienced clinician and one non-clinician) looking for an optimal scanning plane of the four-chamber cardiac view from a fetal phantom. We analysed ultrasound image frames, time series of texture image features and quaternions and found that the experienced clinician performed the procedure in a quicker and smoother way compared to lengthy and non-constant movements from non-clinicians. For future work, we pointed out the need of pruned and quantised neural network models for real-time applications in robotic ultrasound-guidance procedure. The resources to reproduce this work are available at \url{https://github.com/mxochicale/rami-icra2023}.

Figure. (a) Ultrasound-guidance procedures: sonographer operating an ultrasound machine with fetal phantom and sensor fusion signals from inertial sensors and ultrasound imaging; (b) Simple framework for skill transfer learning: collecting experience with sensors ($Pt_n$ pose and $St$ Signal), sampling method for fusion sensor ($\Delta_t$), and identified the need of computational efficient neural network model ($\Omega_\theta$), and output for high-dimensional model \cite{deng2021}, and (c) Robotic ultrasound-guidance procedures: transformations, graphical user interface and simulation using robotic US-guidance light-weight 7 degrees-of-freedom robot (KUKA LBR Med 7) \cite{Gerlach2022, Ipsen2021}.

For the 2023 Robot-Assisted Medical Imaging (ICRA-RAMI) workshop

Important dates:

• Abstract Submission Deadline: 15th March 2023， (extension 24th March 2023)
• Author Notification: 1st April 2023
• Workshop Date: 29th May 2023

See program, accepted papers and further information about the 2023 ICRA-RAMI workshop 🔗.

Clone repository

• Generate your SSH keys as suggested here (or here)
• Clone the repository by typing (or copying) the following line in a terminal at your selected path in your machine:

cd && mkdir -p $HOME/repositories/mxochicale && cd  $HOME/repositories/mxochicale
git clone git@github.com:mxochicale/rami-icra2023.git

Citations

BibTeX to cite

@misc{leung-xochicale-rami-icra2023,
      author={
            Tsz Yan Leung and 
            Miguel Xochicale},
      title={
            Towards a Simple Framework of Skill Transfer Learning for 
            Robotic Ultrasound-guidance Procedures}, 
      year={2023},
      eprint={2305.04004},
      archivePrefix={arXiv},
      primaryClass={cs.RO}
}

Contributors

Thanks goes to all these people (emoji key):

Goosie Leung 💻 🤔 🔧

Miguel Xochicale 💻 🔬 🤔 🔧 📖 🔧

This work follows the all-contributors specification.
Contributions of any kind welcome!