Correlation of endoscopic optical coherence tomography with histology in the lower-GI tract.


Journal Article

BACKGROUND:Optical coherence tomography (OCT), a noninvasive optical imaging technique, provides high-resolution cross-sectional images of tissue microstructure. We developed a system for real-time endoscopic OCT (EOCT) of the human GI tract. During clinical trials, the structure of mucosa and submucosa, glands, blood vessels, pits, villi, and crypts was observed in a range of GI organs. Although EOCT images are thought to accurately depict actual histologic features, there are few data to support this assumption. Therefore, the present study correlated images acquired with an EOCT imaging system in vitro to corresponding histologic sections. METHODS:EOCT images were obtained of fresh specimens of ileum, colon, and rectum that then were fixed in formalin and were processed for microscopic evaluation by using standard methods. The thickness of mucosa and of submucosa was determined for both EOCT images and histologic slides. RESULTS:The first hyper-reflective layer in the EOCT images was identified as mucosa. A close correlation (R 2=0.84) was observed between EOCT and histology. Furthermore, the submucosa and the muscularis propria could be identified as the next deepest hyporeflective band and a hyper-reflective layer, respectively, in EOCT images. The submucosa was found to be more compressible than mucosa, and its EOCT appearance was dependent on its content of adipose tissue. CONCLUSIONS:EOCT provides images that precisely correlate with the histologic structure of the mucosa and the submucosa of the GI tract.

Full Text

Duke Authors

Cited Authors

  • Westphal, V; Rollins, AM; Willis, J; Sivak, MV; Izatt, JA

Published Date

  • April 2005

Published In

Volume / Issue

  • 61 / 4

Start / End Page

  • 537 - 546

PubMed ID

  • 15812406

Pubmed Central ID

  • 15812406

Electronic International Standard Serial Number (EISSN)

  • 1097-6779

International Standard Serial Number (ISSN)

  • 0016-5107

Digital Object Identifier (DOI)

  • 10.1016/s0016-5107(05)00084-2


  • eng