A detailed comparison between the endoscopic images using blue laser imaging and three-dimensional reconstructed pathological images of colonic lesions

Takeshi Ueda; Kohei Morita; Fumikazu Koyama; Yuichi Teramura; Tadashi Nakagawa; Shinji Nakamura; Yayoi Matsumoto; Takashi Inoue; Takayuki Nakamoto; Yoshiyuki Sasaki; Hiroyuki Kuge; Maiko Takeda; Chiho Ohbayashi; Hisao Fujii; Masayuki Sho

doi:10.1371/journal.pone.0235279

A detailed comparison between the endoscopic images using blue laser imaging and three-dimensional reconstructed pathological images of colonic lesions

PLoS One. 2020 Jun 29;15(6):e0235279. doi: 10.1371/journal.pone.0235279. eCollection 2020.

Authors

Takeshi Ueda^{1

2}, Kohei Morita³, Fumikazu Koyama^{1

4}, Yuichi Teramura⁵, Tadashi Nakagawa⁶, Shinji Nakamura⁷, Yayoi Matsumoto¹, Takashi Inoue¹, Takayuki Nakamoto^{1

4}, Yoshiyuki Sasaki¹, Hiroyuki Kuge¹, Maiko Takeda³, Chiho Ohbayashi³, Hisao Fujii⁸, Masayuki Sho¹

Affiliations

¹ Department of Surgery, Nara Medical University, Kashihara, Japan.
² Department of Surgery, Minami-Nara General Medical center, Yoshino, Nara, Japan.
³ Department of Diagnostic Pathology, Nara Medical University, Kashihara, Japan.
⁴ Department of Endoscopy, Nara Medical University Hospital, Kashihara, Japan.
⁵ Clinical Research Endoscopy System Division and Medical System Business Division, FUJIFILM Corporation, Tokyo, Japan.
⁶ Department of Surgery, Saiseikai Chuwa Hospital, Sakurai, Japan.
⁷ Department of Surgery, Takanohara Central Hospital, Nara, Japan.
⁸ Gastrointestinal Endoscopy and IBD center, Yoshida Hospital, Nara, Japan.

Abstract

Blue laser/light imaging (BLI) is an image-enhanced endoscopy (IEE) technique that can provide an accurate diagnosis by closely observing the surface structure of various colonic lesions. However, complete correspondence between endoscopic images and pathological images has not been demonstrated. The aim of this study was to accurately compare endoscopic images and the pathological images using a three-dimensionally (3D) reconstructed pathological model. Continuous thin layer sections were prepared from colonic tissue specimens and immunohistochemically stained for CD34 and CAM5.2. Three-dimensional reconstructed images were created by superimposing immunohistochemically stained pathological images. The endoscopic image with magnifying BLI was compared with the top view of the 3D reconstructed image to identify any one-to-one correspondence between the endoscopic images and histopathological images using the gland orifices and microvessels as a guide. Using 3D reconstructed pathological images, we were able to identify the location on the endoscope image in cases of colonic adenocarcinoma, adenoma and normal mucosa. As a result, the horizontal plane of the endoscopic image and the vertical plane of the 2D pathological specimen were able to be compared, and we successfully determined the visible blood vessel depth and performed a detailed evaluation on magnifying BLI. Examples are as follows: (1) The median vasculature depth from the mucosal surface that could be recognized as vasculature on magnifying BLI was 29.4 μm. The median depth of unrecognizable vessels on magnifying BLI was 218.8 μm, which was significantly deeper than recognizable vessels. (2) Some brownish structures were suggested to potentially be not only dense vessels, vessel expansions, corrupted vessels but also bleeding or extravasation of erythrocytes. Overall, we demonstrated a new approach to matching endoscopic images and pathological findings using a 3D-reconstructed pathological model immunohistochemically stained for CD34 and CAM5.2. This approach may increase the overall understanding of endoscopic images and positively contribute to making more accurate endoscopic diagnoses.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Adenocarcinoma / pathology*
Aged
Colonic Neoplasms / pathology*
Early Detection of Cancer / methods*
Endoscopy / methods*
Female
Humans
Imaging, Three-Dimensional / methods*
Male
Middle Aged
Narrow Band Imaging

Grants and funding

The study was funded by FUJIFILM Corporation. The funder provided support in the form of salaries for authors [Y.T] and financial support in the form of research materials, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section. No additional external funding was received for this study.