Frontiers of robotic endoscopic capsules: a review

Ciuti, Gastone; Caliò, Renato; Camboni, Domenico; Neri, Luca; Bianchi, Federico; Arezzo, Alberto; Koulaouzidis, Anastasios; Schostek, Sebastian; Stoyanov, Danail; Oddo, Calogero Maria; Magnani, Bernardo; Menciassi, Arianna; Morino, Mario; Schurr, M. O.; Dario, P.

doi:10.1007/s12213-016-0087-x

Cited by 124 publications

(77 citation statements)

References 147 publications

(135 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, the 5-year survival rate of CRC patients is lower than 7% when the disease reaches an advanced stage, however, in cases of early diagnosis, it increases to more than 90% with successful treatment. 1 Conventional colonoscopy is considered the reference standard for CRC screening and diagnosis. It provides direct visualization of the inner surface of the colon, acquire biopsies and perform therapeutic procedures on early stage neoplastic lesions.…”

Section: Introductionmentioning

confidence: 99%

“…Despite the popularity of the method, the success of the exam highly depends on the operator skills, as a high level of hand-eye coordination is required to examine the majority of the colon wall. 1 Leufkens et al reported that missed polyp detection rates could reach values as high as 25%. 2 One effective way to increase the detection rate in colonoscopy exams is the incorporation of computer-aided diagnostic systems.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fully convolutional neural networks for polyp segmentation in colonoscopy

et al. 2017

Self Cite

View full text Add to dashboard Cite

Colorectal cancer (CRC) is one of the most common and deadliest forms of cancer, accounting for nearly 10% of all forms of cancer in the world. Even though colonoscopy is considered the most effective method for screening and diagnosis, the success of the procedure is highly dependent on the operator skills and level of hand-eye coordination. In this work, we propose to adapt fully convolution neural networks (FCN), to identify and segment polyps in colonoscopy images. We converted three established networks into a fully convolution architecture and fine-tuned their learned representations to the polyp segmentation task. We validate our framework on the 2015 MICCAI polyp detection challenge dataset, surpassing the state-of-the-art in automated polyp detection. Our method obtained high segmentation accuracy and a detection precision and recall of 73.61% and 86.31%, respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Fully convolutional neural networks for polyp segmentation in colonoscopy

et al. 2017

Self Cite

View full text Add to dashboard Cite

show abstract

“…Wireless capsule endoscopy (WCE) systems constitute an important class of the aforementioned biomedical systems [1]. The interest in WCE has increased with the effect of noticeable gastrointestinal (GI) disease statistics.…”

Section: Introductionmentioning

confidence: 99%

Localization and Tracking of Implantable Biomedical Sensors

Umay

Fi̇dan

Barshan

2017

Sensors

View full text Add to dashboard Cite

Implantable sensor systems are effective tools for biomedical diagnosis, visualization and treatment of various health conditions, attracting the interest of researchers, as well as healthcare practitioners. These systems efficiently and conveniently provide essential data of the body part being diagnosed, such as gastrointestinal (temperature, pH, pressure) parameter values, blood glucose and pressure levels and electrocardiogram data. Such data are first transmitted from the implantable sensor units to an external receiver node or network and then to a central monitoring and control (computer) unit for analysis, diagnosis and/or treatment. Implantable sensor units are typically in the form of mobile microrobotic capsules or implanted stationary (body-fixed) units. In particular, capsule-based systems have attracted significant research interest recently, with a variety of applications, including endoscopy, microsurgery, drug delivery and biopsy. In such implantable sensor systems, one of the most challenging problems is the accurate localization and tracking of the microrobotic sensor unit (e.g., robotic capsule) inside the human body. This article presents a literature review of the existing localization and tracking techniques for robotic implantable sensor systems with their merits and limitations and possible solutions of the proposed localization methods. The article also provides a brief discussion on the connection and cooperation of such techniques with wearable biomedical sensor systems.

show abstract

“…When the disease reaches an advanced stage, the 5-year survival rate of CRC patients is lower than 7%, but early diagnosis with successful treatment can dramatically increase this figure to more than 90%. 1 Conventional colonoscopy is the reference standard for CRC screening and diagnosis. It provides direct visualization of the inner surface of the colon for acquiring biopsies and performing therapeutic procedures on early stage neoplastic lesions.…”

Section: Introductionmentioning

confidence: 99%

Towards a Computed-Aided Diagnosis System in Colonoscopy: Automatic Polyp Segmentation Using Convolution Neural Networks

Brandão

Zisimopoulos

Mazomenos

et al. 2018

J. Med. Robot. Res.

Self Cite

View full text Add to dashboard Cite

Early diagnosis is essential for the successful treatment of bowel cancers including colorectal cancer (CRC) and capsule endoscopic imaging with robotic actuation can be a valuable diagnostic tool when combined with automated image analysis. We present a deep learning rooted detection and segmentation framework for recognizing lesions in colonoscopy and capsule endoscopy images. We restructure established convolution architectures, such as VGG and ResNets, by converting them into fully-connected convolution networks (FCNs), fine-tune them and study their capabilities for polyp segmentation and detection. We additionally use Shapefrom-Shading (SfS) to recover depth and provide a richer representation of the tissue's structure in colonoscopy images. Depth is incorporated into our network models as an additional input channel to the RGB information and we demonstrate that the resulting network yields improved performance. Our networks are tested on publicly available datasets and the most accurate segmentation model achieved a mean segmentation IU of 47.78% and 56.95% on the ETIS-Larib and CVC-Colon datasets, respectively. For polyp detection, the top performing models we propose surpass the current state of the art with detection recalls superior to 90% for all datasets tested. To our knowledge, we present the first work to use FCNs for polyp segmentation in addition to proposing a novel combination of SfS and RGB that boosts performance.

show abstract

Frontiers of robotic endoscopic capsules: a review

Cited by 124 publications

References 147 publications

Fully convolutional neural networks for polyp segmentation in colonoscopy

Fully convolutional neural networks for polyp segmentation in colonoscopy

Localization and Tracking of Implantable Biomedical Sensors

Towards a Computed-Aided Diagnosis System in Colonoscopy: Automatic Polyp Segmentation Using Convolution Neural Networks

Contact Info

Product

Resources

About