Automatic lesion detection in capsule endoscopy based on color saliency

Drs Koulaouzidis and IakovidisDimitris K. Iakovidis, MSc, PhD and Anastasios Koulaouzidis, MD, FRCPE from The Royal Infirmary of Edinburgh & University of Edinburgh (AK), Department of Computer Engineering, Technological Educational Institute of Central Greece in Lamia, Greece report on their article “Automatic lesion detection in capsule endoscopy based on color saliency: closer to an essential adjunct for reviewing software.”

The focus of this study was to assess the validity of innovative, automatic lesion-detection software in wireless capsule endoscopy (WCE).

The advent of WCE has revolutionized the diagnostic approach to small-bowel disease. However, the task of reviewing WCE video sequences is laborious and time-consuming; software tools offering automated video analysis would enable a timelier and potentially a more accurate diagnosis.

A total of 137 deidentified WCE single images, 77 showing pathology and 60 normal images, were used. These images were obtained with MiroCam (IntroMedic Co, Seoul, South Korea). This image dataset is now available via the following link: Based on the capsule endoscopy standardized terminology (CEST), the lesions were classified into 4 categories: vascular, inflammatory, lymphangiectatic, and polypoid. For each abnormality, a graphic annotation was constructed using Ratsnake (also available via the link: Annotated images were exported as masks, ie, black-and-white frames in which any pathology is depicted as white area on a black background. A color feature-based pattern recognition methodology was devised and applied to the aforementioned image group.The proposed methodology, unlike state-of-the-art approaches, is capable of detecting several different types of lesions. The average performance, in terms of the area under the receiver-operating characteristic curve, reached 89.2 ± 0.9%. The best average performance was obtained for angiectasias (97.5 ± 2.4%) and nodular lymphangiectasias (96.3 ± 3.6%).

Figure 6Figure 6. Classification results per pathology obtained by using the proposed methodology. The error bars represent SD. AUC, area under the receiveroperating characteristic curve.

A color feature–based pattern recognition methodology is capable of automatic detection of several different types of lesions. It outperforms previous state-of-the-art approaches and exhibits robustness in the presence of luminal content while it is capable of detecting even the very small lesions.

Find the abstract for this article online here.

The information presented in Endoscopedia reflects the opinions of the authors and does not represent the position of the American Society for Gastrointestinal Endoscopy (ASGE). ASGE expressly disclaims any warranties or guarantees, expressed or implied, and is not liable for damages of any kind in connection with the material, information, or procedures set forth.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s