Introduction

Linked color imaging (LCI), a unique image-enhanced endoscopy technique, expands the color range to enhance the recognition of subtle mucosal differences, facilitating early lesion detection. Recently, the advantage of LCI in cancer visibility and detection has been reported especially in Helicobacter pylori (H. pylori)-negative gastric cancer including cases post-eradication.¹ However, the correlations between the color of LCI’s visualized mucosa and histological findings have not been thoroughly investigated.

Aims & Methods

This study aimed to investigate relationships between color values and histological findings visualized by LCI in H. pylori-negative gastric cancer.
Thirty-two patients with H. pylori-negative, differentiated-type intramucosal gastric cancer (≤20 mm) who underwent endoscopic submucosal dissection were retrospectively reviewed. Images captured with LCI and white light imaging (WLI) were analyzed using the CIE1976 (L^*a^*b^*) color space, as defined by the International Commission on Illumination (CIE).² Three regions of interest, each measuring 10 × 10 pixels, were set on the malignant tumor and adjacent mucosa to the border. The values of a^*(red-green) and b^* (yellow-blue) and the color difference (ΔE), which represents the overall difference between the two colors, were calculated. Histopathological evaluation using Hematoxylin-Eosin stained specimens was performed to highlight the demarcation between cancerous and noncancerous mucosa in a width of 1.0 mm inside and outside the lesion. Vascular density, glandular density, mucosal thickness, and the degrees of intestinal metaplasia (IM) were measured. Vascular density was calculated as the percentage of the occupied blood vessels in an area up to 200 μm from the surface. And the proportion of IM relative to the glands was measured and classified as the severe IM (≥60%), moderate IM (≥30% and <60%), or mild IM (<30%) groups.

Results

LCI significantly increased ΔE between the cancer and surrounding mucosa compared with WLI (P < 0.001); however, no significant differences in Δa^* and Δb^* were observed between LCI and WLI. Moderate correlations were observed between LCI a^* values and vascular density (ρ = 0.39, P < 0.001) and between LCI b^* values and the proportion of IM (ρ = −0.49, P < 0.001). Furthermore, the vascular density of cancer in the 0−100 μm depth showed a moderate correlation with the LCI a^* values (ρ = 0.41, P < 0.001). Severe IM significantly decreased the LCI b^* values compared with mild and moderate IM (P = 0.02, P < 0.001).

Conclusion

LCI quantitatively visualizes the superficial vascularity of cancer lesions and the presence of IM of the background mucosa in H. pylori-negative patients. It is a potential noninvasive tool for gastric cancer risk stratification and early detection.

References

1. Kato M, Ono S, Kawada K et al. Diagnostic performance of linked color imaging for gastric cancer by Helicobacter pylori infection status: a subanalysis of the large-scale, multicenter randomized controlled trial LCI-FIND. Helicobacter 2024; 29: e1308.
2. Kuehni RG. Color-tolerance data and the tentative CIE 1976 L a b formula. J Opt Soc Am 1976; 66: 497–500.

Disclosure

Shoko Ono has received lecture and/or consulting fees from Fujifilm.

Introduction

Endoscopic submucosal dissection (ESD) was developed in Japan decades ago and is now widely used as a minimally invasive treatment in Western countries. On-the-job training is the major method for acquiring ESD skills; thus, learning ESD is difficult for foreign endoscopists, who have less opportunity to perform it in their country. ENDONIX (Olympus, Tokyo, Japan) is a laptop-based next-generation 3D printed endoscopy simulator that allows basic and advanced endoscopy training. This study aims to evaluate the effectiveness of ENDONIX training for improving novice foreign endoscopists’ ESD skills.

Aims & Methods

This prospective, single-center, exploratory study was conducted at National Cancer Center Hospital (NCCH) between January 2024 and January 2025. Foreign endoscopists who had performed fewer than three ESD cases and participated in an international training program at NCCH were included. Participants were assigned 1:1 to either the Group A (Day 1-14: ENDONIX training with ESD observation followed by Day 15-28: ESD observation alone) or Group B (Day 1-14: ESD observation alone followed by Day 15-28: ENDONIX training with ESD observation). During the training period, participants performed three basic and one advanced training program on ENDONIX every day. On Days 1, 15, and 29, the endoscopists performed ESD for a 2cm lesion at the lesser curvature of the middle gastric body using a non-animal ESD model (G-Master, KOTOBUKI MEDICAL, Tokyo, Japan), and their technical performance was assessed at each time point. All procedures were supervised by a single experienced endoscopist (TU). The primary endpoint was the improvement in dissection speed (mm/min²) from Day 1 to Day 15 between the groups. Paired t-test was used to compare the corresponding continuous variables.

Results

		Group A	Group B	P value
Day 1	mean (SD) [95% CI], mm2/min	12.2 (2.8) [6.8 – 17.7]	13.4 (2.2) [9.1 – 17.7]	0.75
Day 15	mean (SD) [95% CI], mm2/min	18.6 (2.8) [13.2 – 24.1]	12.6 (1.4) [9.9 – 15.4]	0.10
Day 29	mean (SD) [95% CI], mm2/min	27.4 (2.8) [22.0 – 32.8]	18.4 (1.6) [15.2 – 21.5]	< 0.05

A total of 10 foreign endoscopists were enrolled in this study. The median age was 34 in Group A and 28 years in Group B. Nationalities in Group A were Spanish (n=3), Portuguese (n=1), and Chinese (n=1); in Group B, Spanish (n=4) and Ukrainian (n=1). All Group A endoscopists specialized in gastroenterology; Group B included four gastroenterologists and one oncologist. No significant differences were found between groups in baseline endoscopic experience. All participants completed predetermined ENDONIX training during the study period. ESD was performed on all designated days using G-Master, and en-bloc resection was achieved for the target lesions on without intraoperative perforation or damaging the resected specimen. The mean dissection speed significantly increased from Day 1 to Day 15 in Group A (“ENDONIX training with ESD observation”) (12.2 vs 18.6, p < 0.05). No significant differences were observed from Day 1 to Day 15 in Group B (“ESD observation alone”) (13.4 vs. 12.6, p = 0.81), while the mean dissection speed significantly increased from Day 15 to Day 29 (12.6 vs. 18.4, p < 0.05). Although the difference on Day 15 was not statistically significant between the groups (18.6 vs. 12.6, p = 0.10), Group A demonstrated a significantly higher dissection speed than Group B on Day 29 (27.4 vs. 18.4 p < 0.05), (Table 1).

Conclusion

The systematic endoscopy training with the ENDONIX platform improved the novice foreign endoscopists’ ESD skills. Dissection speed increased in the period receiving systematic ENDONIX training, highlighting the importance of systematic endoscopy training.

Disclosure

None

Introduction

There is increasing recognition that effective resident training in gastrointestinal endoscopy should be provided by endoscopists who have received formal educational training in the teaching of endoscopy.¹ While the implementation of Training the Colonoscopy Trainers (TCT) courses has led to significant improvements in colonoscopy quality outcomes^2,3 and polyp detection rates,⁴ evidence regarding the impact of TCT participation on teaching performance and resident learning – both of which are main goals of faculty development initiatives – is still lacking. Therefore, the aim of the present study was to evaluate the effectiveness of TCT course implementation using Kirkpatrick’s model of training evaluation.

Aims & Methods

In this study, 24 gastroenterologists from a university hospital and two affiliated general teaching hospitals in the north-eastern educational region of the Netherlands, participated in a two-day TCT course. The course was evaluated at three levels of Kirkpatrick’s model: (a) post-course evaluations reflecting the initial reaction; (b) video analysis of endoscopy training in the workplace before and after the TCT course using the mini-Structured Training Trainer Assessment Report (mini-STTAR); (c) a post-course resident survey reflecting the impact on the home-environment. Descriptive statistics and non-parametric tests were used to analyse the effectiveness of the TCT course.

Results

The TCT course was well received by participants, they highlighted the use of a structured training framework as a strength. Video analysis of endoscopy training in the workplace demonstrated a significant improvement in post-course trainer performance, both at 1-3 months (T1) and 6-9 months (T2). Significant improvements were found in the overall mini-STTAR scores (p<0.001), and across most mini-STTAR items, particularly in the items related to the phase preceding the actual training (referred to as the Set) and the evaluation after completion of the training (referred to as the Closure). Regarding the actual training (referred to as the Dialogue), participants demonstrated a significant positive change in ‘guiding/deconstructing’ and ‘controlling’ (both at T1 and T2) and in ‘encouraging/praise’ (only at T2). Participants received significantly higher 5-point Likert scale ratings as ‘excellent trainer’ at T1 (mean 4.13) and T2 (mean 4.29), compared to baseline (mean 3.58, p<0.001). In the post-course survey, all residents (N=16) were satisfied with the supervision of endoscopy trainers in their teaching hospitals. Nine (56%) reported uniform endoscopy supervision among trainers, a significant increase compared to a pre-course resident survey (8%, p=0.034).⁵ Although residents reported that the frequency of a didactic Set and Closure did not increase, most respondents indicated that the implementation of the TCT course improved the setting of learning objectives, the provision of spontaneous feedback, guidance during difficult endoscopies, and the prevention of undesired scope takeover.

Conclusion

The implementation of a TCT course provided a structured and useful training framework, and enhanced trainer performance, particularly in the didactic pre- and post-endoscopy training discussions, as well as in guiding, controlling, encouraging, and providing spontaneous feedback during the training of residents in gastrointestinal endoscopy.

References

1. Antonelli G, Voiosu AM, Pawlak KM, et al. Training in basic gastrointestinal endoscopic procedures: a European Society of Gastrointestinal Endoscopy (ESGE) and European Society of Gastroenterology and Endoscopy Nurses and Associates (ESGENA) Position Statement. Endoscopy. 2024;56:131-150
2. Waschke KA, Anderson J, Macintosh D et al. Training the gastrointestinal endoscopy trainer. Best Pract Res Clin Gastroenterol. 2016;30:409-419
3. Gavin DR, Valori RM, Anderson JT, et al. The national colonoscopy audit: a nationwide assessment of the quality and safety of colonoscopy in the UK. Gut. 2013;62:242-249
4. Hoff G, Botteri E, Huppertz-Hauss G, et al. The effect of train-the-colonoscopy-trainer course on colonoscopy quality indicators. Endoscopy. 2021;53:1229-1234
5. Mousset RA, de Vos Tot Nederveen Cappel WH, Pierie JEN et al. Endoscopy training in the Netherlands: a national survey among gastroenterology residents. Endosc Int Open. 2022;10:E1548-E1554

Introduction

The incidence of superficial non-ampullary duodenal epithelial tumors (SNADETs) has markedly increased in recent endoscopic practice [1, 2]. Nevertheless, the detection rate during esophagogastroduodenoscopy remains low, with 0.15% for adenomas and 0.05% for carcinomas [3]. Large SNADETs, akin to other gastrointestinal malignancies, should be resected en bloc via endoscopic submucosal dissection (ESD). However, the incidence of perforation associated with duodenal ESD is notably high (11.6–35.7%), often necessitating invasive interventions such as pancreatoduodenectomy [4–6]. Endoscopists must learn duodenal ESD procedures through on-the-job training, as no suitable animal or simulation model exists. Although cadaver surgical training (CST) has become prevalent in surgical practice, its application in gastrointestinal endoscopic treatment has been infrequently documented. CST may provide superior anatomical landmarks and tissue fidelity compared to animal or simulation training. This study investigated the potential application of CST as a training model for duodenal ESD.

Aims & Methods

Between July 2024 and February 2025, CST was conducted on four donors preserved using the Thiel method. Upper gastrointestinal examination (UGIE) and multiple duodenal ESD procedures were performed. Among the participants, three had prior experience with these procedures (Experience; Group E), while four did not have such experience (Non-Experience; Group NE). Both groups executed all procedures, and subsequent retrospective studies were conducted: 1) the feasibility of UGIE and the performance of the duodenal ESD procedure for each donor, and 2) questionnaires administered to the participants (n=7) to assess the feasibility of each procedure (marking, submucosal injection, circumferential incision, submucosal dissection), realism of the simulation, and psychological barriers, using a 5-point rating scale ("1: significantly more challenging–5: significantly easier" or "1: strongly disagree–5: strongly agree"). Each score was compiled and presented as the median score.

Results

1) UGIE was feasible in all four cases, although one instance of duodenal perforation occurred due to duodenal stretching. Duodenal ESD was feasible in all four cases and was performed a total of seven times.
2) In UGIE, procedural scores, such as insertion and observation at each location, were predominantly rated 4 or 5, with a score of 3 exclusively noted for intragastric observation. In duodenal ESD, all procedural scores for tasks such as submucosal injection and dissection were consistently assigned a rating of 4. The questionnaire score for the item "Was the duodenal mural architecture preserved?" was 4.5. The scores for the items "Is there psychological stress associated with handling cadavers?" and "Is CST useful for learning duodenal ESD technique?" were 4 and 4.5, respectively. No significant differences were observed in the scores between the E and NE groups for any of the evaluations.

Conclusion

The mural architecture of the gastrointestinal tract in CST, when fixed using the Theil method, closely resembles that of a living organism. It was subsequently identified that CST could serve as an effective educational model for duodenal ESD, marking this as the inaugural report on the subject. This methodology is expected to have broader applications in the education of endoscopic treatments, particularly for rare and challenging procedures such as duodenal ESD, and to contribute to the advancement of endoscopic techniques in future research.

References

1. Legué LM, Bernards N, Gerritse SL, et al. Trends in incidence, treatment and survival of small bowel adenocarcinomas between 1999 and 2013: a population-based study in The Netherlands. Acta Oncol. 2016;55:1183-1189.
2. Yoshida M, Yabuuchi Y, Kakushima N, et al. The incidence of non-ampullary duodenal cancer in Japan: The first analysis of a national cancer registry. J Gastroenterol Hepatol. 2021;36:1216-1221.
3. Yoshimizu S, Hirasawa T, Horiuchi Y, et al. Differences in upper gastrointestinal neoplasm detection rates based on inspection time and esophagogastroduodenoscopy training. Endosc Int Open. 2018;06:E1190–E1197.
4. Hoteya S, Furuhata T, Takahito T, et al. Endoscopic Submucosal Dissection and Endoscopic Mucosal Resection for Non-Ampullary Superficial Duodenal Tumor. Digestion. 2017;95:36-42.
5. Jung JH, KD. C, JY. A, et al. Endoscopic submucosal dissection for sessile, nonampullary duodenal adenomas. Endoscopy. 2013;45:133-135.
6. Kato M, Takeuchi Y, Hoteya S, et al. Outcomes of endoscopic resection for superficial duodenal tumors: 10 years' experience in 18 Japanese high volume centers. Endoscopy. 2022;54:663-670.

Introduction

Prophylactic wound closure after Endoscopic Submucosal Dissection (ESD) is suggested in high-risk patients to reduce delayed AEs (e.g. bleeding). The closure of large wounds can be challenging with standard endoscopic devices (e.g. endoclips) and dedicated devices are needed. Endoscopic Hand-Suturing (EHS) can be a therapeutic option to close extensive gastrointestinal defects. The current study presents the EHS experience from two Italian centers, aiming to assess the feasibility and clinical effectiveness of the EHS technique.

Aims & Methods

Data from two high-volume endoscopic Italian centers were retrospectively collected. Patients with an ESD wound closed using the SutuArt (Olympus Medical Systems, Tokyo, Japan) device were enrolled in the study (Figure1). Demographics, technical data, success rate, suturing speed, and AEs were recorded

Results

Between June 2023 and August 2024 10 patients were enrolled in the study, demographics in Table 1. All patients were at high-risk for delayed bleeding after gastric or colorectal ESD (figure2). The mean lesion size was 34mm (range 15-45). In all cases SutuArt (Olympus Medical Systems, Tokyo, Japan) needle holder with V-LOC 180, size 3-0 threads (Medtronic) were used. All procedures were performed by two expert endoscopists. ESD-wound closure was successfully achieved in 90% of cases. The mean suturing time was 42 min (range: 25–65 min, SD=23.3 min) meanwhile the mean ESD time was 86 minutes (range: 15–185min, SD=38min). Suturing significantly increased the total procedure time (120min Vs 86min, p<0.001). In one case, the wound closure failed due to suture breakage, resulting in needle entrapment within the mesorectum, with the impossibility of retrieving it. No other AEs were reported.

Conclusion

ESD-defects closure with EHS is feasible and safe with promising outcomes. However, this technique is significantly time-consuming. Probably extended training may be required for endoscopists performing suturing, and further studies are needed to assess the real-life applicability and the cost-effectiveness of this technique.

Introduction

Artificial Intelligence (AI)-powered Computer-aided diagnosis (CADx) systems provide real-time polyp histology prediction during colonoscopy, potentially improving accuracy in clinical decision-making. We aimed to assess CADx performance and its impact on endoscopist optical diagnosis for sessile serrated polyps (SSPs), traditional serrated adenomas (TSAs), adenomas with advanced histology, and cancers in a pragmatic implementation study.

Aims & Methods

We performed a secondary analysis of a large prospective cohort comprising patients undergoing diagnostic, screening, and surveillance colonoscopies. All patients and polyps with SSPs, TSAs, advanced adenomas and/or cancers detected, and prospectively documented CADx assisted or CADx unassisted optical diagnoses were included in the study. The primary outcome was the sensitivity of CADx assisted and unassisted optical diagnoses for SSP identification. Secondary outcomes included sensitivity for TSA, advanced adenoma, and cancer identification as well as the impact of CADx output on endoscopists’ optical diagnoses.

Results

Within a cohort of 2622 patients with 2456 eligible polyps there were 140 (5.7%) SSPs, 22 (0.9%) TSAs, 213 (8.7%) advanced adenomas, and 13 (0.5%) cancers. Sensitivity for SSP identification of CADx assisted and unassisted optical diagnosis was 53.2% (95% CI 42.3-63.8) and 37.8% (95% CI 24.1-53.9) respectively; p=0.12. In CADx assisted cases, endoscopists correctly identified SSPs more frequently when CADx classified these as non-neoplastic (56/79 cases) compared to neoplastic (23/79 cases), (66.1% [95% CI 53.0-77.1] vs 21.7% [95% CI 9.6-41.9]; p=0.0003). Sensitivity for TSA identification as neoplastic by CADx output was 61.6% (8/13), but CADx assisted and unassisted endoscopist optical diagnosis performed similarly, identifying TSAs as neoplastic in 83.3% and 66.7% of cases respectively; p=0.61. Sensitivity for advanced adenoma identification as adenoma/neoplastic did not differ between CADx assisted and unassisted cases (Table 1). Sensitivity for optical diagnosis of cancer was 50.0% and 42.9% for CADx assisted and unassisted optical diagnosis, with 3/6 and 4/7 of cancers misdiagnosed as adenomas, respectively.

Table 1. Diagnostic performance of CADx assisted and unassisted optical diagnosis

	SSP		TSA		Advanced adenomas*		Cancers
	CADx assisted	CADx unassisted	CADx assisted	CADx unassisted	CADx assisted	CADx unassisted	CADx assisted	CADx unassisted
Polyps, n	79	37	12	9	71	83	6	7
Sensitivity, % (95% CI)	53.2 (42.3-63.8)	37.8 (24.1-53.9)	83.3 (54.0-97.0)	66.7 (34.7-88.4)	100.0 (95.0-100)	96.4 (89.3-98.9)	50.0 (11.8-88.2)	42.9 (9.9-81.6)
PPV, %	27.8	20.0	-	-	-	-	23.1	33.3
Incorrect endoscopist diagnoses
Hyperplastic, n	17	4	1	2	0	0	0	0
Adenoma, n	20	19	-	-	-	-	3	4
SSP, n	-	-	1	1	0	0	0	0
Cancer, n	0	0	0	0	0	3	-	-

*Adenomas with villous component or high grade dysplasia

Conclusion

In our study, CADx use did neither harm nor improve optical diagnostic performance in polyps with SSP, TSA, advanced adenoma, or cancerous histology. Sensitivity for identification of these lesions trended universally higher with CADx assisted optical diagnosis, though statistical significance was not achieved at the present sample size. Whilst both CADx assisted and unassisted optical diagnosis reliably identified advanced adenomas as adenomas, optical diagnosis sensitivity to recognize SSPs and cancerous lesions was low.

Disclosure

Daniel von Renteln has received research funding from ERBE Elektromedizin GmbH, Ventage, Pendopharm, Fujifilm, and Pentax, and has received consultant or speaker fees from Boston Scientific Inc., ERBE Elektromedizin GmbH, and Pendopharm. The remaining authors declare that they have no conflict of interest.

Introduction

Training with silicone models in colonoscopy is limited by the lack of realistic texture and appearance of mucosal tissue. Artificial intelligence (AI) offers a novel solution by enabling simulation of realistic appearing mucosa, which can enhance the overall visual realism of silicone-based models for endoscopic training.

Aims & Methods

We developed, using 239,875 images from 3,452 colonoscopies recorded using five different endoscopic processors, a style transfer AI model capable of transforming endoscopic video footage from a silicone colon into a video that visually mimics real mucosal tissue. The model was applied to two representative clips recorded from a silicone colon model, covering lesions and normal mucosa. Both the original and AI-enhanced videos were shown to N = 12 gastroenterologists (board certified and in training), who provided quantitative evaluations using a 5-point Likert scale (1: Strongly disagree, 5: Strongly agree) and free-text qualitative feedback. We conducted the statistical analysis using Wilcoxon signed-rank tests with Pratt adjustment for zero differences and analyzed the qualitative feedback with an inductive content analysis after Mayring.

Results

All physicians prefer the AI-enhanced videos over the original silicone footage due to the more realistic representation of the colon. This can also be seen from the quantitative data, as the participants rated the AI-enhanced videos as significantly more visually realistic (M = 2.42 vs 4.17, Z = -3.22, p < .001), helpful for training (M = 3.67 vs 4.25, Z = -2.65, p = .016) and polyp characterization (M = 2.92 vs 3.75, Z = -2.38, p = .027), favored for usage (M = 2.92 vs 4.00, Z = -2.80, p = .005), and facilitating optical diagnoses (M = 3.58 vs 4.42, Z = -2.23, p = .039). Though, the participants rated the original silicone footage as significantly less mentally demanding than the AI-enhanced videos (M = 1.67 vs 2.42, Z = -2.61, p = .016). Open-ended feedback highlighted the more realistic AI generated mucosa as an enhancement (100 %) to the unrealistic silicone surface (92 %) while suggesting that further development should focus on integrating more realistic lesions (33 %) and increasing mucosal details (42 %).

Conclusion

AI-driven simulation of normal mucosa improves the realism, acceptance, and educational value of silicone colon models used in endoscopy training.