Deep EndoVO: A recurrent convolutional neural network (RCNN) based visual odometry approach for endoscopic capsule robots

Abstract

Ingestible wireless capsule endoscopy is an emerging minimally invasive diagnostic technology for inspection of the GI tract and diagnosis of a wide range of diseases and pathologies. Medical device companies and many research groups have recently made substantial progresses in converting passive capsule endoscopes to active capsule robots, enabling more accurate, precise, and intuitive detection of the location and size of the diseased areas. Since a reliable real time pose estimation functionality is crucial for actively controlled endoscopic capsule robots, in this study, we propose a monocular visual odometry (VO) method for endoscopic capsule robot operations. Our method lies on the application of the deep Recurrent Convolutional Neural Networks (RCNNs) for the visual odometry task, where Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) are used for the feature extraction and inference of dynamics across the frames, respectively. Detailed analyses and evaluations made on a real pig stomach dataset proves that our system achieves high translational and rotational accuracies for different types of endoscopic capsule robot trajectories.

GANVO: Unsupervised Deep Monocular Visual Odometry and Depth Estimation with Generative Adversarial Networks

“…To solve this, we apply GANs that provide shaper and more accurate depth maps. The second issue of the aforementioned unsupervised techniques is the fact that they only employ CNNs that only analyse just-in-moment information to estimate camera pose [5], [7]. We address this issue by employing a CNN-RNN architecture to capture temporal relations across frames.…”

Unsupervised Odometry and Depth Learning for Endoscopic Capsule Robots

“…As an emerging example, various diseases such as colorectal cancer and inflamatory bowel disease are diagnosed by the usage of swallowable capsule endoscopes, which are noninvasive, painless, suitable to be used for long duration screening purposes which can access difficult body parts (e.g.,small intestines) better than standard endoscopy. Such benefits make swallowable, non-tethered capsule endoscopes an exciting alternative over standard endoscopy [1], [2].…”

RMDL

“…There is a need to develop more general information processing methods for classification and categorization across a broad range of data types. While many researchers have successfully used deep learning for classification problems (e.g., see [9,23,28,30,51]), the central problem remains as to which deep learning architecture (DNN, CNN, or RNN) and structure (how many nodes (units) and hidden layers) is more efficient for different types of data and applications. The favored approach to this problem is trial and error for the specific application and dataset.…”