Correspondence: Shinwa Tanaka Division of Gastroenterology, Department of Internal Medicine, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe, Hyogo 650-0017, Japan E-mail: tanakas@med.kobe-u.ac.jp

Received 2024 May 13; Revised 2024 July 30; Accepted 2024 August 8.

Abstract

Background/Aims

Visualization of palisade vessels (PVs) in Barrett’s esophagus is crucial for proper assessment. This study aimed to determine whether red dichromatic imaging (RDI) improves PV visibility compared with white-light imaging (WLI) and narrow-band imaging (NBI).

Methods

Five expert and trainee endoscopists evaluated the PV visibility in Barrett’s esophagus using WLI, NBI, and RDI on 66 images from 22 patients. Visibility was rated on a 4-point scale: 4, excellent; 3, good; 2, fair; and 1, poor. The color difference between the most recognizable PV spots and surrounding areas with undetectable blood vessels was also analyzed.

Results

Mean visibility scores were 2.6±0.7, 2.3±0.6, and 3.4±0.4 for WLI, NBI, and RDI, respectively. The RDI scores were significantly higher than the WLI (p<0.001) and NBI (p<0.001) scores. These differences were recognized by trainees and expert endoscopists. Color differences in PVs were 7.74±4.96 (WLI), 10.43±5.09 (NBI), and 15.1±6.54 (RDI). The difference in RDI was significantly higher than that in WLI (p<0.001) and NBI (p=0.006).

Conclusions

RDI significantly improved PV visibility compared to WLI and NBI based on objective and subjective measures.

Keywords: Barrett esophagus; Color difference; Palisade vessels; Red dichromatic imaging; Visibility

Graphical abstract

INTRODUCTION

Chronic gastroesophageal reflux disease causes reflux esophagitis (RE) and increases the risk of Barrett’s esophagus. Barrett’s adenocarcinoma can develop within the mucosa of Barrett’s esophagus. With the westernization of lifestyle in Japan, there are concerns about potential increases in the incidence of Barrett’s esophagus and adenocarcinoma.1 Recent reports, however, indicate a less pronounced increase in the incidence of Barrett's adenocarcinoma compared with that in Western countries.2

The accurate identification of the gastroesophageal junction (GOJ) is necessary for the diagnosis of Barrett’s esophagus. Several endoscopic landmarks have been used to define the GOJ, such as the proximal end of the gastric fold (PEGF) and the distal end of the palisade vessels (DEPVs).3,4 However, both definitions have limitations. The position of PEGF can vary with the amount of air insufflation and can be obscured in severe gastric atrophy. Conversely, DEPV visualization can be challenging unless the GOJ is fully distended during deep inspiration, or in cases of surrounding mucosal inflammation.

In 2022, an international consensus report on the GOJ was published, emphasizing that DEPVs provide a more suitable option than PEGF for GOJ definition based on anatomical evidence.5 Therefore, optimal visualization of the palisade vessels (PVs) is essential for the accurate diagnosis of Barrett’s esophagus.

Recently, various image-enhanced endoscopy (IEE) techniques, including narrow-band imaging (NBI) and blue laser imaging, have become integral to clinical practice. Olympus Corp. recently developed a new-generation endoscopic system, the EVIS X1. This system features a novel IEE technology called red dichromatic imaging (RDI), which can enhance the visualization of deep blood vessels in the mucosa.6 This feature suggests potential improvements in PV visualization compared to conventional white-light imaging (WLI) and NBI. This study aimed to investigate whether RDI provides superior visibility of PVs compared to WLI and NBI.

METHODS

Patients

This retrospective observational study used prospectively collected endoscopic images obtained between November 2022 and June 2023 at the Kobe University Hospital and Tanaka Clinic. Patients scheduled to undergo esophagogastroduodenoscopy (EGD) using the EVIS X1 system (Olympus Corp.) and GIF-XZ1200 (Olympus Corp.) were considered for inclusion in this study. Indications for endoscopy included heartburn, abdominal pain, decreased appetite, and anemia. Endoscopy was not limited to patients with specific symptoms or diseases. Inclusion criteria were as follows: confirmed Barrett’s esophagus visible with WLI, NBI, and RDI, and age ≥18 years. The exclusion criteria were as follows: age <18 years, prior gastrectomy or esophageal surgery history, and lack of clear squamocolumnar junction (SCJ) visualization using WLI, RDI, and NBI. Opt-out methods were used to obtain informed consent.

Study protocol

All EGDs were performed by expert endoscopists who had performed more than 10,000 procedures. Endoscopic images were obtained at similar distances and angles by using WLI, NBI, and RDI to ensure a circumferential view of the SCJ. The GOJ was defined as the PEGF, with the area between the GOJ and SCJ diagnosed as Barrett’s esophagus using WLI. The primary outcome measured was the visibility score obtained using WLI, RDI, and NBI. Secondary outcomes included color differences assessed using WLI, RDI, and NBI; visibility scores based on endoscopic experience; and the correlation between visibility scores and color differences. Barrett’s esophagus was classified as long-segment Barrett’s esophagus (LSBE) if it was more than 3 cm circumferentially and short-segment Barrett’s esophagus (SSBE) if less than 3 cm circumferentially or non-circumferential.7 Hiatus hernia was assessed using Hill's flap valve grade.8 Atrophic gastritis was classified on the basis of the Kimura and Takemoto classification system as either open-type (O: 1–3) or closed-type (C: 1–3).9 RE was classified according to the modified Los Angeles classification system.10,11 Patients who had never been tested for Helicobacter pylori infection at the time of endoscopy were tested for H. pylori. However, the patients who indicated in the pre-endoscopy interview that they had a history of H. pylori eradication were not retested. H. pylori infection status was determined by the presence of anti-H. pylori antibodies or ¹³C-urea breath test results. The cutoff values for H. pylori antibodies were as follows: positive, 10 U/mL; high-titer negative, 3 to 9.9 U/mL; and negative, <3 U/mL. In patients with serum levels between 3 and 9.9 U/mL, the ¹³C-urea breath test was performed as an additional test.

RDI

RDI, also known as dual-red imaging, is designed to enhance the visualization of deep vessels and bleeding points. It uses specific wavelengths of green (520–540 nm), amber (595–610 nm), and red (620–640 nm) light. These wavelengths were selected because of their minimal light-scattering characteristics compared with NBI. The latter two wavelengths penetrate deep into the mucosa, making deep vessels more prominent. RDI offers three observational modes: modes 1 and 2 prioritize large vessel visibility in deep tissue, whereas mode 3 enhances both deep large vessels and shallower mucosal vessels based on how each monochromatic image is assigned to RGB color channels.6 In this study, mode 3 was used for all RDI endoscopic images.

Evaluation of the visibility score

Ten endoscopists (five experts and five trainees) evaluated the visibility of PVs within Barrett’s esophagus in 66 images from 22 patients. The scoring system was as follows: 4 (excellent, easily detectable); 3 (good, detectable with careful observation); 2 (fair, difficult to detect without careful examination); and 1 (poor, not detectable without repeated careful examination). Each endoscopist scored each image independently (Fig. 1). Experts were defined as endoscopists who had performed >5, 000 procedures, whereas trainees performed <1,000 procedures. WLI, NBI, and RDI scores were compared for each modality.

Fig. 1.

Sample images of visibility scores. Arrows indicate palisade vessels within Barrett’s esophagus. WLI, white-light imaging; NBI, narrow-band imaging; RDI, red dichromatic imaging.

Evaluation of color difference

We evaluated the color difference between 20×20-pixel spots where PVs were the most recognizable and the surrounding areas where blood vessels were undetectable within Barrett’s esophagus (Fig. 2). The color differences in the images were analyzed and compared with those in the WLI, RDI, and NBI modes. Color differences were assessed by comparing the color values of the regions of interest (ROI) using Adobe Photoshop Elements 2020 (Adobe Systems Inc.).

Fig. 2.

Representative endoscopic images captured using (A) white-light imaging (WLI), (B) narrow-band imaging (NBI), and (C) red dichromatic imaging (RDI). The color difference (ΔE^∗) of the pixel values based on L^∗a^∗b^∗ color spaces within the region of interest (ROI) was analyzed to evaluate the visibility of palisade vessels (PVs). Red lines delineate ROIs (20×20 pixels). PVs are the most recognizable. Yellow lines delineate ROIs surrounding areas where blood vessels are undetectable within Barrett’s esophagus.

Objective evaluation: Lab color space

The CIELAB color space system L*a*b* (L*=light/dark, a*=red/green, b*=yellow/blue) color scores were used for objective evaluation.12 Color values (L, a, and b) in the ROI and the average values were calculated using a histogram panel. L, a, and b represent the color scores in Photoshop (lab color units). These values were converted to L*a*b* color values using the formula: L*=L/256×100, a*=a−128, b*=b–128.13,14 For L*a*b* color spaces in a given ROI, color differences (ΔE*=[(ΔL^*)²+(Δa^*)²+(Δb^*)²]^1/2) in pixel values were used to assess the visibility of each color image.

Correlation between color difference and visibility score

To assess the contribution of color contrast to PV recognition in Barrett’s esophagus, we evaluated the correlation between color difference and visibility scores for all endoscopists, both experts and trainees.

Statistical analysis

Wilcoxon rank-sum test was used to assess statistically significant differences in (1) visibility scores rated by trainees and experts and in (2) ΔE* and L*a*b* color values between images. Statistical significance was set at a p-value of <0.05. The degree of correlation between the color difference and visibility scores was evaluated using Spearman’s correlation coefficient, in which values of <0.3 denoted a weak correlation, 0.3 to 0.7 showed a moderate correlation, and >0.7 denoted a strong correlation.15 These calculations were performed using R statistical software ver. 4.3.1 (The R Foundation for Statistical Computing).

A preliminary study was conducted prior to this study. The mean visibility score (±standard deviation) for RDI was 3.60±0.28, and that for WLI was 2.87±0.56, with a correlation coefficient of –0.47 between both groups. Considering that the effectiveness of the RDI was comparable with that of previous studies, the effect size was estimated to be 0.99. Thus, using a lower threshold for relative Pitman efficiency of 0.864 in comparing the Wilcoxon test and t-test, a total sample size of 18 patients with α=0.05 and β=0.05 (power=0.95) was estimated. Anticipating a dropout rate of 10 %, a final sample of 20 patients was included. The sample size was calculated using power analysis software (G*Power 3; Universität Düsseldorf).

Ethical statements

The study protocol was reviewed and approved by the Institutional Review Board of Kobe University Hospital (approval number: B230029). This study conformed to the principles of the Declaration of Helsinki. Given the retrospective nature of the study, informed consent was waived, and the participants were provided with information about the study and given the right to opt out.

RESULTS

Participants

Between November 2022 and June 2023, 22 patients underwent EGD using WLI, RDI, or NBI. Patient characteristics are summarized in Table 1. The median patient age was 55 years. Of the 22 participants, 14 were male, and eight were female. All patients underwent SSBE, and none underwent LSBE. Hiatal hernia grading using Hill’s classification system revealed eight patients classified as grade I, eight patients as grade II, and six patients as grade III. RE was graded according to the modified Los Angeles classification system: 10 patients were classified as grade N, three as grade M, seven as grade A, and two as grade B. Four patients were administered proton pump inhibitors. Atrophic gastritis was observed in eight patients, all of whom had closed-type gastritis. H. pylori infection was present in three patients, eradicated in five patients, and unknown in one patient; 15 patients were uninfected.

Table 1.

Baseline characteristics of patients

Visibility score

Table 2 shows the visibility scores for WLI, NBI, and RDI as assessed by all endoscopists, both trainees and experts. The mean visibility scores across all endoscopists were 2.6±0.7, 2.3±0.6, and 3.4±0.4 in WLI, NBI, and RDI, respectively. The RDI scores were significantly higher than both WLI (p<0.001) and NBI (p<0.001) scores in all the cases. These differences were observed in both trainees and expert endoscopists. Furthermore, the WLI scores of the experts were significantly higher than the NBI scores (p=0.007).

Table 2.

Evaluation of WLI, NBI, and RDI modes for visibility scores

Table 3 shows the visibility scores of each endoscopist on WLI, NBI, and RDI. The RDI scores were consistently the highest among all endoscopists. In addition, no significant differences were found between the trainee and expert visibility scores for each modality.

Table 3.

Comparison of visibility scores between trainees and experts

Color difference

Table 4 shows calculations for L*a*b* color values for PVs and surrounding areas where blood vessels were undetectable within Barrett’s esophagus region. Significant differences were observed in the following: (1) a* and b* color difference values between WLI and NBI; (2) L*, a*, and b* color difference values between NBI and RDI; and (3) L* and a* color difference values between WLI and RDI. Color differences (ΔE^∗) of PVs were 7.74±4.96 in WLI, 10.43±5.09 in NBI, and 15.1±6.54 in RDI. The difference (ΔE^∗) in the RDI mode was significantly higher than that in WLI (p<0.001) and NBI modes (p=0.006). In addition, the difference (ΔE^∗) in the NBI mode was significantly higher than that in the WLI mode (p=0.017).

Table 4.

Objective evaluations using color differences (ΔL^*, Δa^*, Δb^*, ΔE^*) in WLI, NBI, and RDI modes

Correlation between visibility score and color difference

Figure 3 illustrates the correlation between the visibility score and the color difference across all modalities. A moderate correlation was observed between the color difference and the visibility score for all endoscopists (γ=0.55, p<0.001), experts (γ=0.54, p<0.001), and trainees (γ=0.56, p<0.001).

Fig. 3.

Correlation between color difference and visibility score in all modes. WLI, white-light imaging; NBI, narrow-band imaging; and RDI, red dichromatic imaging.

DISCUSSION

This study investigated the effectiveness of RDI in enhancing the visibility of PV collaterals compared with the effectiveness of WLI and NBI. Both objective and subjective assessments demonstrated significantly improved PV visibility with the RDI. In addition, the visibility scores for RDI were higher than those for the other modes, regardless of the endoscopic experience. This is the first study to report that RDI offers superior visualization of PVs compared to WLI and NBI.

Although patients with LSBE are considered to be at high risk for Barrett's adenocarcinoma, the prevalence of LSBE is lower in Japan than in Western countries, with SSBE representing the majority of cases.16 However, there have been reports of Barrett's adenocarcinoma originating from SSBE,17,18 highlighting the importance of accurate diagnosis encompassing both LSBE and SSBE. Accurate diagnosis of Barrett’s esophagus also plays a pivotal role in identifying esophagogastric junction cancers. By assessing the background of the lesion, it becomes feasible to predict whether the lesion arises from Barrett’s esophagus or the cardiac region. Several landmarks have been used to define the GOJ, including the SCJ, PEGF, and DEPV.19 The DEPV is considered the most preferred landmark as it is unaffected by epithelial changes and have been validated by anatomical, histological, and in vivo anatomic studies.20-23 Therefore, a new modality that improves PV visibility is warranted.

Recently, various IEEs have been reported to improve the visibility of superficial structures and vessels more effectively than WLI.24-26 Several studies on the visibility of Barrett’s esophagus have been published. Hamamoto et al.27 reported that NBI improves the diagnostic ability of Barrett’s esophagus compared to WLI. However, their study relied on magnified findings and did not evaluate PV visibility. In contrast, another study that evaluated PV visibility showed no significant differences between NBI and WLI.28 Similarly, blue laser imaging, which functions similarly to NBI, showed no significant difference over WLI in PV visualization.29 While these IEEs significantly enhanced the visibility of blood vessels in the mucosal surface layer, such as in the intrapapillary capillary loop, PVs, located deeper within the lamina propria, remain poorly visualized. Recently, linked color imaging (LCI), a new type of IEE, has demonstrated improved visibility of Barrett’s esophagus compared to WLI. LCI has a specific color-enhancing function that amplifies reddish and whitish hues. LCI is believed to improve visibility by rendering PVs a more distinct purple-red color, even when PVs are difficult to discern with WLI.29 In this study, the visibility of PVs for RDI was significantly improved compared with that of NBI and WLI. The RDI uses green light to depict blood vessels within the superficial and middle mucosal layers, including the PVs. In RDI mode 3, green light is assigned to the green signal that has the highest human visual sensitivity on the monitor displays. This likely contributed to the improvement in PV visibility. However, in RDI modes 1 and 2, the green light was assigned to a blue signal, resulting in lower recognizability and poorer PV visualization.

Several studies have reported the clinical benefits of RDI. These applications include identifying bleeding points, diagnosing and treating esophageal varices, and evaluating ulcerative colitis activity.30-34 The most common studies on this topic are related to the identification of bleeding points. Several investigators have reported that the RDI facilitates hemostasis during ESD.30,31 As for esophageal varices, Furuichi et al.32,33 reported that RDI improved the visibility of esophageal varices, especially those located in the shallow layers. RDI also increased the success rate of endoscopic injection sclerotherapy and decreased variceal recurrence rates. In ulcerative colitis, Naganuma et al.34 developed an RDI-based scoring system to assess disease activity. This score was reported to correlate closely with endoscopic and histologic grades, potentially aiding patient prognosis. Our study is the first to explore the utility of RDI in PV diagnosis.

Analysis of color differences revealed significantly higher values for RDI than for WLI and NBI, mirroring the trend observed in the visibility scores. When comparing NBI and WLI, the color difference was higher for NBI; however, there was no significant difference in the visibility scores among all endoscopists. Despite some discrepancies between the subjective visibility scores and objective color differences, a moderate correlation was observed between these measures.

This study had some limitations. First, it is based on images obtained from a relatively small cohort of patients. Second, the diagnosis of Barrett’s esophagus was based on endoscopic findings without histological confirmation, and only PV visibility was evaluated. Third, all patients had SSBE, with no cases of LSBE included in the patient cohort. Therefore, larger prospective studies combined with histological examinations are necessary to demonstrate the usefulness of RDI more reliably.

In conclusion, this study demonstrates that RDI significantly improves the visibility of PVs compared to WLI and NBI, as evidenced by both objective and subjective assessments. Additionally, the RDI scores were consistently higher than those for the other modalities, regardless of the endoscopist’s experience.

Notes

Conflicts of Interest

The medical equipment used in this study (the endoscope system [EVIS X1] and endoscope [GIF-XZ1200]) were loaned by Olympus Corporation for use at the Kobe University Hospital and Tanaka Clinic during the study period. The authors have no potential conflicts of interest.

Funding

None.

Acknowledgments

This study was supported by three endoscopy evaluation hospitals: Hyogo Prefectural Harima-Himeji General Medical Center (Ryusuke Ariyoshi), Osaka Saiseikai Nakatsu Hospital (Tomoya Sako), and Hyogo Prefectural Tamba Medical Center (Kazunori Tsuda).

Author Contributions

Conceptualization: ST; Data curation: CU; Formal analysis: CU, HA, HS, MK, HTak, HH, RI, SH, HTan, EN, TY, TTa, MT, TTo, YM, YK; Methodology: ST; Validation: all authors; Visualization: all authors; Writing–original draft: CU, HA, HS, MK, HTak, HH, RI, SH, HTan, EN; Writing–review & editing: ST, TY, TTa, MT, TTo, YM, YK.

References

1. Kusano C, Gotoda T, Khor CJ, et al. Changing trends in the proportion of adenocarcinoma of the esophagogastric junction in a large tertiary referral center in Japan. J Gastroenterol Hepatol 2008;23:1662–1665.

2. Matsuno K, Ishihara R, Ohmori M, et al. Time trends in the incidence of esophageal adenocarcinoma, gastric adenocarcinoma, and superficial esophagogastric junction adenocarcinoma. J Gastroenterol 2019;54:784–791.

3. Kuwano H, Nishimura Y, Oyama T, et al. Guidelines for diagnosis and treatment of carcinoma of the esophagus april 2012 edited by the Japan Esophageal Society. Esophagus 2015;12:1–30.

4. Sharma P, Dent J, Armstrong D, et al. The development and validation of an endoscopic grading system for Barrett's esophagus: the Prague C & M criteria. Gastroenterology 2006;131:1392–1399.

5. Sugano K, Spechler SJ, El-Omar EM, et al. Kyoto international consensus report on anatomy, pathophysiology and clinical significance of the gastro-oesophageal junction. Gut 2022;71:1488–1514.

6. Uraoka T, Igarashi M. Development and clinical usefulness of a unique red dichromatic imaging technology in gastrointestinal endoscopy: a narrative review. Therap Adv Gastroenterol 2022;15:17562848221118302.

7. Eluri S, Shaheen NJ. Barrett's esophagus: diagnosis and management. Gastrointest Endosc 2017;85:889–903.

8. Hill LD, Kozarek RA, Kraemer SJ, et al. The gastroesophageal flap valve: in vitro and in vivo observations. Gastrointest Endosc 1996;44:541–547.

9. Kimura K, Takemoto T. An endoscopic recognition of the atrophic border and its significance in chronic gastritis. Endoscopy 1969;1:87–97.

10. Hongo M. Minimal changes in reflux esophagitis: red ones and white ones. J Gastroenterol 2006;41:95–99.

11. Hoshihara Y, Hashimoto M. Endoscopic classification of reflux esophagitis. Nihon Rinsho 2000;58:1808–1812.

12. Kuehni RG. Color-tolerance data and the tentative CIE 1976 L a b formula. J Opt Soc Am 1976;66:497–500.

13. Sommers M, Beacham B, Baker R, et al. Intra- and inter-rater reliability of digital image analysis for skin color measurement. Skin Res Technol 2013;19:484–491.

14. Tomie A, Dohi O, Yagi N, et al. Blue laser imaging-bright improves endoscopic recognition of superficial esophageal squamous cell carcinoma. Gastroenterol Res Pract 2016;2016:6140854.

15. Schober P, Boer C, Schwarte LA. Correlation coefficients: appropriate use and interpretation. Anesth Analg 2018;126:1763–1768.

16. Chang CY, Cook MB, Lee YC, et al. Current status of Barrett's esophagus research in Asia. J Gastroenterol Hepatol 2011;26:240–246.

17. Goda K, Singh R, Oda I, et al. Current status of endoscopic diagnosis and treatment of superficial Barrett's adenocarcinoma in Asia-Pacific region. Dig Endosc 2013;25 Suppl 2:146–150.

18. Matsui A, Kuribayashi Y, Nomura K, et al. Conventional white light endoscopic features of small superficial Barrett’s esophageal adenocarcinoma. Digestion 2016;93:47–52.

19. Boyce HW. Endoscopic definitions of esophagogastric junction regional anatomy. Gastrointest Endosc 2000;51:586–592.

20. Vianna A, Hayes PC, Moscoso G, et al. Normal venous circulation of the gastroesophageal junction. A route to understanding varices. Gastroenterology 1987;93:876–889.

21. Sato T, Kato Y. Palisading longitudinal esophagus vessels at esophago-gastric junction. Hepatogastroenterology 2008;55:305–307.

22. Sato T, Kato Y, Matsuura M, et al. Significance of palisading longitudinal esophagus vessels: identification of the true esophagogastric junction has histopathological and oncological considerations. Dig Dis Sci 2010;55:3095–3101.

23. Aida J, Vieth M, Ell C, et al. Palisade vessels as a new histologic marker of esophageal origin in ER specimens from columnar-lined esophagus. Am J Surg Pathol 2011;35:1140–1145.

24. Oyama T, Monma K. A new classification of magnified endoscopy for superficial esophageal squamous cell carcinoma. Esophagus 2011;8:247–251.

25. Ezoe Y, Muto M, Horimatsu T, et al. Magnifying narrow-band imaging versus magnifying white-light imaging for the differential diagnosis of gastric small depressive lesions: a prospective study. Gastrointest Endosc 2010;71:477–484.

26. Sano Y, Tanaka S, Kudo SE, et al. Narrow-band imaging (NBI) magnifying endoscopic classification of colorectal tumors proposed by the Japan NBI Expert Team. Dig Endosc 2016;28:526–533.

27. Hamamoto Y, Endo T, Nosho K, et al. Usefulness of narrow-band imaging endoscopy for diagnosis of Barrett's esophagus. J Gastroenterol 2004;39:14–20.

28. Schölvinck DW, Goto O, Seldenrijk CA, et al. Detection of palisade vessels as a landmark for Barrett's esophagus in a Western population. J Gastroenterol 2016;51:682–690.

29. Takeda T, Nagahara A, Ishizuka K, et al. Improved visibility of Barrett’s esophagus with linked color imaging: inter- and intra-rater reliability and quantitative analysis. Digestion 2018;97:183–194.

30. Yorita N, Oka S, Tanaka S, et al. Clinical usefulness of dual red imaging in gastric endoscopic submucosal dissection: a pilot study. Clin Endosc 2020;53:54–59.

31. Fujimoto A, Saito Y, Abe S, et al. Haemostasis treatment using dual red imaging during endoscopic submucosal dissection: a multicentre, open-label, randomised controlled trial. BMJ Open Gastroenterol 2019;6e000275.

32. Furuichi Y, Gotoda T, Moriyasu F, et al. Dual red imaging (novel advanced endoscopy) can increase visibility and can predict the depth in diagnosing esophageal varices. J Gastroenterol 2017;52:568–576.

33. Furuichi Y, Gotoda T, Kasai Y, et al. Role of dual red imaging to guide intravariceal sclerotherapy injection of esophageal varices (with videos). Gastrointest Endosc 2018;87:360–369.

34. Naganuma M, Yahagi N, Bessho R, et al. Evaluation of the severity of ulcerative colitis using endoscopic dual red imaging targeting deep vessels. Endosc Int Open 2017;5:E76–E82.

Article information Continued

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Characteristic	Total (n=22)
Age (yr)	55 (31–84)
Male, sex	14 (63.6)
SSBE/LSBE	22/0
Hiatus hernia (Hill classification)
I/II/III	8/8/6
Reflux esophagitis
N/M/A/B	10/3/7/2
Anti-acid therapy	4 (18.2)
Atrophic gastritis
C0/C1–3/O1–3	14/8/0
Helicobacter pylori infection status
Currently infected/uninfected/eradicated/unknown	3/13/5/1

	Trainees (n=5)						Experts (n=5)						p-value Trainees vs. experts
	A	B	C	D	E	All trainees	A	B	C	D	E	All experts	p-value Trainees vs. experts
WLI	2.9±0.9	2.3±1.1	2.3±0.9	2.3±1.2	2.3±0.8	2.4±0.8	2.5±1.0	2.6±0.8	2.3±1.0	3.6±0.6	2.4±0.7	2.7±0.7	0.190
NBI	2.3±0.6	2.3±1.1	2.3±0.6	2.1±1.1	2.3±0.8	2.3±0.7	2.7±0.8	2.3±0.5	1.9±0.8	2.8±0.4	2.0±0.7	2.3±0.5	0.962
RDI	3.2±0.7	3.7±0.5	2.4±1.0	3.0±0.9	3.2±0.6	3.3±0.5	3.9±0.3	3.5±0.6	3.2±0.6	3.7±0.4	2.9±0.7	3.4±0.4	0.507

	WLI	NBI	RDI	p-value (WLI vs. NBI)	p-value (NBI vs. RDI)	p-value (WLI vs. RDI)
ΔL*	4.0±3.4	5.7±6.1	12.1±7.0	0.205	0.001	0.001
Δa*	–4.7±5.1	4.7±6.2	–1.5±7.4	<0.001	0.003	0.048
Δb*	–0.8±2.9	1.8±2.0	–1.3±4.4	0.005	0.005	0.672
ΔE*	7.7±5.0	10.4±5.1	15.1±6.5	0.017	0.006	<0.001

	WLI	NBI	RDI	p-value (WLI vs. NBI)	p-value (NBI vs. RDI)	p-value (WLI vs. RDI)
All	2.6±0.7	2.3±0.6	3.4±0.4	0.295	<0.001	<0.001
Trainee	2.4±0.8	2.3±0.7	3.3±0.5	0.434	<0.001	<0.001
Expert	2.7±0.7	2.3±0.5	3.4±0.4	0.007	<0.001	<0.001