[Factors influencing the miss rate of polyps in a tandem colonoscopy study]

Comparative study of convolutional neural network architectures for gastrointestinal lesions classification

The gastrointestinal (GI) tract can be affected by different diseases or lesions such as esophagitis, ulcers, hemorrhoids, and polyps, among others. Some of them can be precursors of cancer such as polyps. Endoscopy is the standard procedure for the detection of these lesions. The main drawback of this procedure is that the diagnosis depends on the expertise of the doctor. This means that some important findings may be missed. In recent years, this problem has been addressed by deep learning (DL) techniques. Endoscopic studies use digital images. The most widely used DL technique for image processing is the convolutional neural network (CNN) due to its high accuracy for modeling complex phenomena. There are different CNNs that are characterized by their architecture. In this article, four architectures are compared: AlexNet, DenseNet-201, Inception-v3, and ResNet-101. To determine which architecture best classifies GI tract lesions, a set of metrics; accuracy, precision, sensitivity, specificity, F1-score, and area under the curve (AUC) were used. These architectures were trained and tested on the HyperKvasir dataset. From this dataset, a total of 6,792 images corresponding to 10 findings were used. A transfer learning approach and a data augmentation technique were applied. The best performing architecture was DenseNet-201, whose results were: 97.11% of accuracy, 96.3% sensitivity, 99.67% specificity, and 95% AUC.

Key factors for improving adenoma detection rate

INTRODUCTION

Colonoscopy is a fundamental tool in colorectal cancer (CRC) prevention. Nevertheless, one-fourth of colorectal neoplasms are still missed during colonoscopy, potentially being the main reason for post-colonoscopy colorectal cancer (PCCRC). Adenoma detection rate (ADR) is currently known as the best quality indicator correlating with PCCRC incidence.

AREAS COVERED

We performed a literature review in order to summarize evidences investigating key factors affecting ADR: endoscopists education and training, patient management, endoscopic techniques, improved navigation (exposition defect), and enhanced lesions recognition (vision defect) were considered.

EXPERT OPINION

'Traditional' factors, such as split dose bowel preparation, adequate withdrawal time, and right colon second view, held a significant impact on ADR. Several devices and technologies have been developed to promote high-quality colonoscopy, however artificial intelligence may be considered the most promising tool for ADR improvement, provided that endoscopists education and recording are guaranteed.

Adenoma miss rate of polypectomy-referring hospitals is high in Korea

BACKGROUND/AIMS

We evaluated the miss rates of polyps, adenomas, and advanced neoplasia of polypectomy-referring hospitals and risk factors for missed adenomas.

METHODS

We compared medical records and electronic images of initial colonoscopies from polypectomy-referring hospitals with those of corresponding therapeutic colonoscopies from Seoul St. Mary's Hospital obtained from May 2014 to February 2016.

RESULTS

A total of 147 patients (56.6 ± 12.1 years, 37 females) were included. The mean number of polyps and adenomas detected on initial colonoscopy was 2.4 ± 1.7 and 1.7 ± 1.4, respectively. The mean number of additionally detected polyps and adenomas per patient during therapeutic colonoscopy was 1.4 ± 1.8 and 1.0 ± 1.5, respectively. Pooled miss rate for polyps, adenomas, and advanced neoplasia was 36%, 37%, and 11%, respectively. Pooled miss rate for adenomas was significantly higher for right-sided, non-pedunculated, and small (< 1 cm) adenomas (p = 0.031, p = 0.000, and p = 0.000, respectively). The miss rate of polyps, adenomas, and advanced neoplasia per patient was 60%, 49%, and 7%, respectively. Multivariate analysis revealed age and number of adenoma on initial colonoscopy were significantly related with risk for adenoma-missing (p = 0.005 and p = 0.023, respectively).

CONCLUSION

Among patients referred for polypectomy, adenoma is missed in one of two patients and advanced neoplasm is missed in one of 13. Patients with advanced age or multiple adenoma on initial colonoscopy have a higher possibility of missed adenoma. Total colon exploration should be performed carefully during therapeutic colonoscopy.

Tubular Locomotion and Positioning Using Tip Eversion for Endoscopy

Colonoscopy is a minimally invasive procedure to examine the large intestine using a flexible endoscope. Currently, colonoscopic procedures require physically pushing the endoscope through the large intestine which potentially imparts damaging forces on the intestinal wall, requires a high level of expertise to perform safely, and can require extensive procedure time. This paper presents the use of inverted tubular element locomotion (ITEL) for improved endoscope translation and positioning that works via tip eversion of inverted plastic tubes. Experiments were performed to examine the fluid pressure required for locomotion through a large intestinal model and in both straight and curved paths and determine the optimal tube dimensions for insertion via ITEL. Experiments were then performed to compare intestinal forces during manual and ITEL-based insertion. The experiments established a relationship between the tip eversion pressure and the tube thickness and diameter. In addition, it was found that pressure required for locomotion was only minimally impeded by being enclosed in the intestinal manikin but significantly impacted by the curvature of the turn. Measured insertion forces for ITEL were found to be 10.1% less compared to traditional insertion. These results will aid in the future design and development of ITEL for endoscopic positioning.

Differences with experienced nurse assistance during colonoscopy in detecting polyp and adenoma: a randomized clinical trial

OBJECTIVES

This study aims to evaluate whether the participation of an experienced endoscopy nurse in colonoscopy increases the polyp detection rate (PDR) and adenoma detection rate (ADR) of experienced colonoscopists.

METHODS

This study was a randomized controlled trial. Patients were randomly assigned to the experienced colonoscopist alone (single observer) group, or experienced nurse participation (dual observer) group. The primary outcome was the PDR and ADR. The advanced lesion detection rate was also recorded.

RESULTS

A total of 587 patients were included in the analysis. Among these patients, 291 patients were assigned to the single observer group, while 296 patients were assigned to the dual observer group. The PDR was 33% in the single observer group and 41.9% in the dual observer group (P = 0.026), while the ADR was 23.0% in the single observer group and 30.4% in the dual observer group (P = 0.043). No significant difference was found for advanced lesions between groups.

CONCLUSIONS

The present data demonstrated that experienced nurse observation during colonoscopy can improve polyp and adenoma detection rates, even if the colonoscopist is experienced.

TRIAL REGISTRATION

Clinicaltrials.gov No. NCT02292563. https://clinicaltrials.gov/ct2/results?cond=&term=NCT02292563&cntry=&state=&city=&dist=.

Risk factors of missed colorectal lesions after colonoscopy

Several studies have reported a significant rate of missed colorectal polyps during colonoscopy. This study aimed to determine the variables that affect the miss rate of colorectal polyps.We performed a retrospective observational study of patients who, between January 2007 and December 2014, had undergone a second colonoscopy within 6 months of their first. In all patients, the first colonoscopy constituted a screening or surveillance colonoscopy as part of a health check-up, and the patients were referred to the endoscopic clinic if there were meaningful polyps. The miss rate of colorectal polyps was evaluated, as were the variables related to these missed lesions.Among 659 patients (535 men), the miss rate of colorectal polyps was 17.24% (372/2158 polyps), and 38.69% of patients (255/659 patients) had at least 1 missed polyp. The most common site for missed polyps was the ascending colon (29.8%), followed by the sigmoid colon (27.8%). The miss rate of polyps was higher in men [odds ratio (OR) = 1.611, 95% confidence interval (95% CI) = 1.024-2.536], patients with multiple polyps at their first colonoscopy (OR = 1.463, 95% CI = 0.992-2.157), and patients who had a history of polyps (OR = 23.783, 95% CI = 3.079-183.694). Multiple missed polyps were more frequently located in the right colon (OR = 2.605, 95% CI = 1.458-4.657), and the risk of sessile serrated adenoma/polyp was greater in the right colon (OR = 10.418, 95% CI = 2.073-52.353).Endoscopists should pay careful attention in patients who have multiple polyps and in those who have a history of polyps, because such patients are at a high risk of missed polyps in colonoscopy.

Computer vision and augmented reality in gastrointestinal endoscopy

Augmented reality (AR) is an environment-enhancing technology, widely applied in the computer sciences, which has only recently begun to permeate the medical field. Gastrointestinal endoscopy—which relies on the integration of high-definition video data with pathologic correlates—requires endoscopists to assimilate and process a tremendous amount of data in real time. We believe that AR is well positioned to provide computer-guided assistance with a wide variety of endoscopic applications, beginning with polyp detection. In this article, we review the principles of AR, describe its potential integration into an endoscopy set-up, and envisage a series of novel uses. With close collaboration between physicians and computer scientists, AR promises to contribute significant improvements to the field of endoscopy.