Future developments in endoscopic imaging

On-the-fly detection of changes on and below the surface in epithelium mucosal tissue architecture from scattered light

In this paper we present a technique to raise a flag on the fly when a transition occurs between different mucosal architectures on or below the surface. The segmentation is based on a novel difference metric for detecting an abrupt change in the parameters extracted from a Stochastic Decomposition Method (SDM) that models the scattered light reflected from the mucosal tissue structure over an area (2-D scan) illuminated by an optical sensor (fiber) emitting light at either one wavelength or with white light. This work has the potential to enhance the endoscopist's ability to locate and identify abnormal mucosal architectures in particular when the disease is developing below the surface and hence becoming hidden during colonoscopy or endoscopic examination. It also has also potential in helping deciding as to when and where to take biopsies; steps that should lead to improvement in the diagnostic yield.

Can we see epithelium tissue structure below the surface using an optical probe?

This paper answers the question of whether it is possible to detect changes below the surface in epithelium layered structures using a Stochastic Decomposition Method (SDM) that models the scattered light reflected from the layered structure over an area (2-D scan) illuminated by an optical sensor (fibre) emitting light at either one wavelength or with white light. Our technique correlates the differential changes in the reflected tissue texture with the morphological and physical changes that occur in the tissue occurring inside the structure. This work has great potential for detecting changes in mucosal structures and may lead to enhanced endoscopy when the disease is developing to the outside of the mucosal structure and hence becoming hidden during colonoscopy or endoscopic examination. Tests are performed on layered tissue phantoms, and the results obtained show great effectiveness of the model and method in picking up changes in the morphology of the layered tissue phantoms occurring below the surface. We also establish the robustness of the model to changes in viewing depth by testing it on phantoms viewed at different depths. We show that the model is robust to within a 4-mm-deep viewing range.

Non-erosive reflux disease, functional heartburn and gastroesophageal reflux disease; insights into pathophysiology and clinical presentation

Non-erosive reflux disease (NERD) has emerged as a real entity in the spectrum of gastroesophageal reflux disease (GERD) and may, indeed, represent the most common manifestation of reflux disease. Recent information indicates NERD can be further subclassified based on the results of pH studies and suggests that functional heartburn needs to be differentiated as that subgroup of patients in which there appears to be no relationship between symptoms, however 'typical' of GERD, and acid exposure. Other aspects of NERD need to be appreciated, including overlap with functional dyspepsia and the potential variations in response to such therapeutic interventions as acid-suppressive therapy and fundoplication. It seems quite possible that our failure to separate functional heartburn from NERD, in general, has contributed, in large measure, to diagnostic difficulties and therapeutic disappointment in GERD. An appreciation of the range of entities, of different etiology and pathology, within NERD will foster the development of effective approaches to the assessment and therapy of this challenging disorder.

Do 40% of patients resected for barrett esophagus with high-grade dysplasia have unsuspected adenocarcinoma?

Results of studies conducted in the last 2 decades suggest that the detection of high-grade dysplasia in patients with Barrett esophagus is the harbinger of a synchronous adenocarcinoma, which remains undetected even by rigorous biopsy protocols but is discovered during resection of the esophagus. The reported prevalence of synchronous carcinomas ranges from 0% to 75%. Other researchers maintain that appropriate surveillance programs can be used to detect carcinomas at a curable stage and to prevent unnecessary esophagectomies. Both logistical difficulties and potential methodological pitfalls have plagued many studies designed to investigate this issue. A large multicenter study that would stratify participants for hitherto unexplored variables (eg, age, gender, and ethnic background) may be required before the 40% occult cancer prevalence can be either confirmed or refuted. However, the large scale needed for such a study to provide reliable data and new developments in endoscopic imaging (eg, magnification endoscopy and optical coherence tomography) and endoscopic therapy (eg, mucosectomy) are likely to make such a study both ethically unacceptable and logistically and financially unfeasible. Future research should utilize the combination of new endoscopic technologies with the continuing search for validated biomarkers that help predict the biological behavior of Barrett epithelium in individual patients, with a particular focus on the possible development of preneoplastic and neoplastic lesions. Pathologists who chose to shift their focus from the traditional morphological investigation of dysplasia to the search for usable biomarkers can position themselves at the center of innovative research projects that could radically modify the management of patients with Barrett esophagus.

Direct visualization of intestinal villi by high-resolution magnifying upper endoscopy: a validation study

BACKGROUND

New generation videoendoscopes potentially may visualize duodenal villi. This study compared endoscopic findings with this type of instrument to the histopathologic evaluation of duodenal villi.

METHODS

A total of 191 patients underwent upper endoscopy for the purpose of obtaining duodenal biopsy specimens. The findings were assessed independently by 3 experienced observers by using a commercially available, high-resolution, high-magnifying (x2) videoendoscope. The duodenal villous profile was determined by endoscopic magnification and by endoscopic magnification after filling the duodenum with water. With both endoscopic magnification and endoscopic magnification after filling the duodenum with water, villous patterns were scored as the following: definitely present, partially present, or definitely absent. Villous patterns also were histopathologically scored as the following: normal, partial villous pattern, or total villous atrophy.

RESULTS

Interobserver variability was excellent (kappa = 0.93). The concordance between either endoscopic magnification or endoscopic magnification after filling the duodenum with water and histology was 100% for presence/absence of villi. The sensitivity, the specificity, and the positive and negative predictive values of endoscopic magnification for detection of any villous abnormality were 95%, 99%, 95%, and 99%, respectively; the respective values of endoscopic magnification after filling the duodenum with water were 95%, 98%, 92%, and 99%.

CONCLUSIONS

High-resolution magnifying upper endoscopy can reliably predict the presence or the absence of duodenal villi.

Individual variations in mucosa and total wall thickness in the stomach and rectum assessed via endoscopic ultrasound

Endoscopic ultrasound is a unique tool to acquire in vivo data on alimentary tract wall thicknesses of sufficient resolution needed in radiation dosimetry studies. Through their different echo texture and intensity, five layers of differing echo patterns for superficial mucosa, deep mucosa, submucosa, muscularis externa and serosa/adventitia exist within the walls of organs composing the alimentary tract. In this study, retrospective image analyses of patient video data were made for ten examinations of the stomach and eight examinations of the rectum covering a range of patient ages. Thicknesses for stomach mucosa ranged from 1030 +/- 130 microm to 1640 +/- 80 microm (total stomach wall thicknesses from 2.80 +/- 0.12 to 4.23 +/- 0.03 mm). Measurements made for the rectal images revealed rectal mucosal thicknesses from 660 +/- 50 microm to 1130 +/- 250 microm (total rectal wall thicknesses from 2.28 +/- 0.05 to 3.55 +/- 0.43 mm). The mucosa accounted for approximately 32 +/- 7% and approximately 32 +/- 8% of the total thickness of the stomach and rectal wall, respectively. These values can thus be utilized to investigate uncertainties in alimentary tract dosimetry that are based upon fixed reference individual definitions of organ wall structure.

Advances in video and imaging in ureteroscopy

Advances in image processing and display technologies, such as digital imaging, HDTV, and virtual reality, will ultimately allow integration of endoscopic imaging with diagnosis and therapy during ureteroscopic procedures. Further improvements in simulation technology and telemedicine should improve surgical training and greatly benefit patient care.