Physical Intelligence Article 2017

Six Degree-of-Freedom Localization of Endoscopic Capsule Robots using Recurrent Neural Networks embedded into a Convolutional Neural Network

Thumb ticker sm mehmet
Physical Intelligence
  • Postdoctoral Researcher
no image
Physical Intelligence
Thumb ticker sm metin eth vertical small
Physical Intelligence
Guest Researcher
Image toc

Since its development, ingestible wireless endoscopy is considered to be a painless diagnostic method to detect a number of diseases inside GI tract. Medical related engineering companies have made significant improvements in this technology in last decade; however, some major limitations still residue. Localization of the next generation steerable endoscopic capsule robot in six degreeof-freedom (DoF) and active motion control are some of these limitations. The significance of localization capability concerns with the doctors correct diagnosis of the disease area. This paper presents a very robust 6-DoF localization method based on supervised training of an architecture consisting of recurrent networks (RNN) embedded into a convolutional neural network (CNN) to make use of both just-in-moment information obtained by CNN and correlative information across frames obtained by RNN. To our knowledge, our idea of embedding RNNs into a CNN architecture is for the first time proposed in literature. The experimental results show that the proposed RNN-in-CNN architecture performs very well for endoscopic capsule robot localization in cases vignetting, reflection distortions, noise, sudden camera movements and lack of distinguishable features.

Author(s): Turan, Mehmet and Abdullah, Abdullah and Jamiruddin, Redhwan and Araujo, Helder and Konukoglu, Ender and Sitti, Metin
Journal: arXiv preprint arXiv:1705.06196
Year: 2017
Month: May
Day: 17
Bibtex Type: Article (article)
DOI: arXiv:1705.06196
Electronic Archiving: grant_archive

BibTex

@article{turan2017six,
  title = {Six Degree-of-Freedom Localization of Endoscopic Capsule Robots using Recurrent Neural Networks embedded into a Convolutional Neural Network},
  journal = {arXiv preprint arXiv:1705.06196},
  abstract = {Since its development, ingestible wireless endoscopy is considered to be a painless diagnostic method to detect a number of diseases inside GI tract. Medical related engineering companies have made significant improvements in this technology in last decade; however, some major limitations still residue. Localization of the next generation steerable endoscopic capsule robot in six degreeof-freedom (DoF) and active motion control are some of these limitations. The significance of localization capability concerns with the doctors correct diagnosis of the disease area. This paper presents a very robust 6-DoF localization method based on supervised training of an architecture consisting of recurrent networks (RNN) embedded into a convolutional neural network (CNN) to make use of both just-in-moment information obtained by CNN and correlative information across frames obtained by RNN. To our knowledge, our idea of embedding RNNs into a CNN architecture is for the first time proposed in literature. The experimental results show that the proposed RNN-in-CNN architecture performs very well for endoscopic capsule robot localization in cases vignetting, reflection distortions, noise, sudden camera movements and lack of distinguishable features.},
  month = may,
  year = {2017},
  slug = {turan2017six},
  author = {Turan, Mehmet and Abdullah, Abdullah and Jamiruddin, Redhwan and Araujo, Helder and Konukoglu, Ender and Sitti, Metin},
  month_numeric = {5}
}