Introduction

Accurate real-time differentiation of colorectal polyps during colonoscopy is critical for optimal decision-making and cost-effective care. Artificial intelligence (AI)-based tools such as MagentiQ-Eye Colo have been developed to assist endoscopists in predicting polyp histology based on endoscopic features. However, real-world validation against histopathology remains limited.

Aims & Methods

We conducted a retrospective study using colonoscopy data collected at our endoscopy unit. Histologically confirmed colorectal polyps were included, with subtypes classified as adenomas (LGD and HGD), sessile serrated adenomas (SSA), or hyperplastic polyps (HP). The CADe AI tool MagentiQ-Eye Colo was applied retrospectively to stored colonoscopy video recordings to generate polyp histology predictions ("Neoplastic" or "Non-Neoplastic"). These were compared to the corresponding histopathological diagnoses. Diagnostic performance metrics including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated.

Results

A total of 257 polyps were analyzed. Of these, 198 polyps received a conclusive AI-based prediction, while 59 polyps were labeled as "Inconclusive" and were excluded from performance analysis.
Overall, MagentiQ-Q achieved a sensitivity of 96.8% and specificity of 22.3% in detecting neoplastic polyps, with an overall accuracy of 58.1%.
A focused subgroup analysis was performed on 133 polyps classified as adenomas, SSA, or HP:
MagentiQ-Eye achieved:

• 96.2% sensitivity and 3.8% specificity for adenomas (adenoma villosum, adenoma tubulare),
• 70.8% sensitivity and 29.2% specificity for sessile serrated adenomas (SSA),
• 43.3% sensitivity and 56.7% specificity for hyperplastic polyps (HP).

Further analysis of adenoma subtypes showed:

• LGD (n=50): 96.0% predicted as neoplastic.
• HGD (n=29): 97.4% predicted as neoplastic.

Size-based analysis revealed that MagentiQ was more likely to classify polyps as neoplastic with increasing estimated size: in the "Advanced" size category, 98.7% of polyps were predicted as neoplastic, in the "Small" category, this rate was 84.0%, in the "Diminutive" category, the neoplasia prediction rate dropped to 61.5%.

Conclusion

The MagentiQ-Eye Colo AI system shows excellent sensitivity in identifying neoplastic polyps, especially conventional adenomas including both LGD and HGD. However, overprediction of neoplasia was frequent in non-neoplastic lesions, notably hyperplastic polyps and a subset of SSA cases, leading to reduced specificity. Moreover, the system yielded inconclusive predictions in 23% of cases, suggesting a need for algorithm refinement and prospective validation to support its clinical integration.

Introduction

In recent years, artificial intelligence (AI) has been widely explored to enhance capsule endoscopy (CE), with the goal of improving the efficiency of the reading process. While most AI models have been developed for small bowel and colon analysis, the development of esophagogastric (E-G) models has been limited due to the scarcity of frames captured during the procedure, making it challenging to develop a robust model. This study aims to develop an interoperable ubiquitous model capable detecting pleomorphic lesions in the E-G tract.

Aims & Methods

We included 59,482 E-G frames, from 774 CE procedures of 5 centers, to develop a Convolutional Neural Network (CNN). The dataset was divided following an exam-based split, with 90% allocated for training – including a 5-fold cross validation – while the remaining was used for testing. The primary outcomes were: sensitivity, specificity, accuracy and area under the curve (AUC).

Results

During training, the CNN achieved mean sensitivity of 85.0% (IC95% 76.5–93.5), specificity of 96.3% (IC95% 93.9–98.7), accuracy of 93.6% (IC95% 91.7–95.4), and AUC-ROC of 0.98 (IC95% 0.97–0.98). During testing, the sensitivity, specificity and accuracy of CNN were 92.2%, 95.1% and 94.6%, respectively.

Conclusion

CE-AI models capable of assessing E-G are crucial step toward developing a truly robust panendoscopic model. This ubiquitous and interoperable model capable of lesion detection in both esophagus and stomach, achieved good overall accuracy. However, prospective real-world studies comparing its performance with standard upper endoscopy are still needed to validate its clinical applicability.

Introduction

Following sphincter-preserving surgery for rectal cancer, many patients develop low anterior resection syndrome (LARS). Symptoms of LARS include fecal incontinence, urgency, frequent bowel movements, and incomplete evacuation. Treatment of these symptoms through anorectal biofeedback that focuses solely on anal sphincter function has shown limited benefits (Koifman et al.). This may be due to a relatively preserved sphincter function in these patients but a disruption in the anorectal apparatus as a working unit. Therefore, we hypothesize that a comprehensive anorectal biofeedback (cBF) program, which addresses both anal sphincter weakness and issues of recto-anal incoordination and rectal sensory abnormalities, may provide greater benefits for patients with LARS.

Aims & Methods

We aimed to evaluate the effectiveness of cBF in LARS patients. Consecutive LARS patients referred a tertiary referral centre were prospectively enrolled. Surgical clearance and rectal examination were required to rule out local complications (e.g., strictures), and anorectal manometry was performed in all patients. A physiotherapist, guided by a gastroenterologist, performed 4-6 cBF session using instrumental anal EMG and a rectal balloon (guided by ARM results and limited to 100 ml). Symptom severity, LARS scores and quality-of-life questionnaires were collected pre- and post-therapy, and at 6-month follow-up. Primary endpoints were end of treatment Clinical Global Impression (CGI) and improvement in LARS scores.

Results

22 patients (median age 59, IQR 53-69; 68% male) were included. Median time from ileostomy closure was 12 months (IQR 6.0-33.0). Median baseline LARS score was 36 (IQR 34-41). Most patients completed 4 cBF sessions (73%), the remainder completed 6 (27%). Baseline ARM showed median maximum resting pressure of 91 mmHg (IQR 77-115) and squeeze pressure of 200 mmHg (IQR 139-255), indicating relatively preserved sphincter function. Post-cBF, the median LARS score improved to 29 (IQR 19-39), with a median within-patient change of -11 (IQR -19 to -6; p=0.002). CGI ratings indicated 73% of patients as "improved," 18% "minimally improved," and 9% reported "no change."

Conclusion

cBF appears effective for LARS, at least in the short term. Addressing recto-anal incoordination and sensory abnormalities may provide additional benefit, even in patients with preserved sphincter function. These findings suggest integrating physician supervised ARM-guided cBF into multidisciplinary care for post-LAR patients may be of importance.

References

1. Koifman E, et al. Long-term persistence and risk factors for anorectal symptoms following low anterior resection for rectal cancer. BMC Gastroenterol. 2024;24(1):31.

Introduction

Several Randomized Controlled Trials (RCTs) have evaluated the role of artificial intelligence (AI) Computer-Aided Detection (CADe) systems to enhance polyp detection. However, these studies have focused exclusively on locally deployed hardware-based AI systems, whereas cloud-based systems may improve the accessibility and upgrading of CADe technology. ‘CADDIE’ (Odin Vision, an Olympus company) is a cloud-based CADe system that receives the live endoscopy image stream via the hospital’s internet network, where the AI algorithm processes the images, and overlays real-time bounding boxes on the endoscopy screen to highlight suspected polyps.

Aims & Methods

This prospective RCT aimed to evaluate the efficacy and safety of the novel cloud-based CADe-system in improving polyp detection.

This international, multi-centre, parallel-group RCT was conducted across eight centres in four European countries. Participating endoscopists had a baseline of >1000 colonoscopy procedures and an adenoma detection rate (ADR) of ≥25%. Patients aged ≥40 years and at average risk for colorectal cancer undergoing screening or surveillance indications for colonoscopy were recruited. Patients were randomised in a 1:1 ratio to standard colonoscopy (SoC) or CADDIE-assisted (CADe-arm) colonoscopy. The co-primary endpoints were superiority in adenomas per colonoscopy (APC) as the efficacy endpoint, and non-inferiority for the Positive Percent Agreement (PPA), defined as the percentage of resections that are histologically confirmed adenomas, SSLs and large (>10mm) hyperplastic polyps of the proximal colon, as the safety endpoint.

Results

Overall, 22 endoscopists and 841 participants were included in the analysis. The CADe-arm significantly increased APC by 33% (0.62 vs 0.82, Ratio 1.33 [95%CI 1.06-1.67], p=0.01) and ADR (35.9% vs 43.2%, difference of 7.3% [95%CI 0.7%-13.9%], p=0.03) compared to the SoC-arm. The mean detection of sessile serrated lesions (SSLs) (0.03 vs 0.08, Ratio 3.30 [95%CI 1.41-7.57], p<0.01), large adenomas (0.04 vs 0.09; Ratio 1.93 [95%CI 1.03-3.62], p=0.04), and large polyps (0.05 vs 0.12; Ratio 2.36 [95%CI 1.33-4.17], p<0.01) also significantly improved. A higher proportion of large polyps (27.3% vs. 37.0%) and large adenomas (15.8% vs 24.2%) were non-polypoid ‘flat’ morphology in the CADe-arm compared to the SoC-arm. Other detection metrics, including neoplastic polyps per colonoscopy (0.66 vs 0.91; Ratio 1.39 [95%CI 1.12-1.73], p<0.01) and polyps per colonoscopy (1.06 vs 1.41 (Ratio 1.35 [95%CI 1.13-1.61], p<0.01), also significantly increased.

Non-inferiority in PPA was achieved (53.4% vs 53.9%; 0.5% (-5.0%,∞)), with no increase in unnecessary resections.

Outcome	Study Arm	Patients (N)	Results	Difference (95% CI)	P-value
APC	SoC	424	0.62 ± 1.19	1	0.01
	CADe	417	0.82 ± 1.40	1.33 (1.06, 1.67)
ADR	SoC	424	35.9%	0	0.03
	CADe	417	43.2%	7.3% (0.7%, 13.9%)
SSL-PC	SoC	424	0.03 ± 0.16	1	<0.01
	CADe	417	0.08 ± 0.45	3.30 (1.41, 7.57)
PPC	SoC	424	1.06 ± 1.53	1	<0.01
	CADe	417	1.41 ± 1.85	1.35 (1.13, 1.61)

Table 1: Study results. ADR = Adenoma detection rate, APC = mean Adenomas Per Colonoscopy, PPC = Polyps Per Colonoscopy, SSL-PC = Sessile Serrated Lesions Per Colonoscopy.

Conclusion

This trial demonstrates the safety and efficacy of a cloud-based CADe-system to increase endoscopists' polyp detection across multiple metrics, notably for large adenomas, large polyps, and SSLs.

Disclosure

AR, CS, LBL, RK have received medical consultancy fees from Odin Medical Ltd

Introduction

High-resolution colonic manometry (HRCM) is the gold standard for measuring colonic contractility and guiding decisions for subtotal colectomy in refractory slow-transit constipation (STC). However, it is only available in a handful of centers worldwide. Body surface colonic mapping (BSCM) is a recently validated technology that allows for easy-to-apply and non-invasive transcutaneous evaluation of pre- and post-prandial colonic motility, and has shown strong correlation with HRCM.

Aims & Methods

We sought to evaluate pre- and post-prandial colonic activity using BSCM in patients with refractory STC and correlate with real-time symptoms.
We applied BSCM to refractory chronic constipation patients while undergoing standard diagnostic HRCM. After bowel prep and an overnight fast a HRCM catheter (36 sensors, 3 cm spacing) was advanced and clipped to the caecum. Two BSCM electrode arrays (each 8x8 sensors, 2 cm spacing, Alimetry) were placed on the abdomen after skin prep. Simultaneous HRCM and BSCM recordings were obtained during a 1-hour fasting period, 1 hour after a standardized low-fat meal, and 1 hour after luminally applied 10 mg of bisacodyl. BSCM electrical recordings were band-pass filtered to isolate colonic cyclic motor patterns (CMPs) of 4-10 cycles per minute. The BSCM motility index (MI) was computed from the global continuous wavelet transform spectrum and compared to a real-time symptom score (1-10 scale). Meal response was quantified based on the fasted-fed ratio (FFR) and response time (RT), a marked rise in MI relative to mealtime. Symptoms were recorded for 15 min. during 3,5h.

Results

A total of 9 patients with refractory STC (mean age 45, range 27-60, 22% male) were enrolled. BSCM MI correlated with bloating in 6 subjects, whereof strongly in 2 subjects (R= 0.53 and 0.74). Abdominal discomfort was even correlated to the BSCM MI in 8 subjects, whereof strongly in 2 (R0.74 and 0.79). Flatulence was only reported by 6 subjects and correlated to the BCSM MI in 4 of them, of which 2 strongly (Rs=0.67). While the desire to defecate was correlated with BSCM MI in 6 subjects, the urge to defecate was in 5. However, in both symptoms, there was a strong correlation in 3 subjects (R between 0.69-0.85). These symptoms were mainly provoked by the administration of bisacodyl, followed by strong colonic contractions. Lastly, BSCM MI correlated well with stomach ache symptoms in 6 subjects, of which 2 strongly (R=0.74 and 0.84). Overall symptom burden was strongly correlated with BSCM MI in 3 subjects (R between 0.61-0.8). One subject had clear increases in symptoms, while the colonic activity was low, whereas another subject had an almost perfect correlation between the two. The other subjects had an overall average correlation between symptoms and colonic motility.

Conclusion

BSCM is an easy-to-apply, painless, non-invasive tool that has the potential to elucidate underlying mechanisms driving symptoms in patients with refractory STC. Relating colonic motility to symptoms could aid in developing a more personalized approach to manage these patients' clinical treatment pathway

Introduction

The treatment of esophageal motor disorders (EMD) has historically been focused on decreasing pressure at the gastro-esophageal junction (GEJ) by disrupting muscular fibers at this level. The classic management through dilatation, botulinum toxin injection, and Heller's myotomy (HM) underwent a revolution with the advent of third-space endoscopy, offering safe access to the submucosal space and enabling peroral endoscopic myotomy (POEM).
In this prospective national multicenter study, the results of the implementation period of this technique in Spain from 2016 to the present are analyzed, examining its short, medium, and long-term outcomes. Furthermore, outcomes based on EMD, previous treatments, and other multiple variables are also investigated.

Aims & Methods

The primary aim of this study was to ascertain the clinical efficacy of POEM.
Secondary objectives included the identification of predictors of clinical success and the measurement of the incidence of post-POEM GERD and significant complications, as well as the determination of the risk factors for complications
This is a prospective multicenter study conducted across 13 hospitals in Spain. Symptomatic patients undergoing POEM for any EMD were recruited between June 2016 and February 2023. Baseline characteristics, procedural details and outcomes related to efficacy and safety during short and long-term follow-up were analyzed.

Results

A total of 441 patients were enrolled, with type II achalasia representing the most common diagnosis (62.1%). POEM was the primary treatment in 277 patients (62.8%). Clinical success data were available for 396 patients, with a success rate of 94.4% at the last available follow-up (2 years in 117 patients, 1 year in 140, <1 year in 139). A higher baseline Eckardt score was an independent risk factor for clinical failure in the long-term follow-up (p = 0.024). Symptomatic reflux occurred in 4.7% of the patients, with a significant esophagitis rate of 10.1%. Complication incidence was 9.3% (41/441), including 14 (3.2%) severe events. Older age (p = 0.025) and total procedure duration (p=0.036) were independent risk factors for complications, while mucosal injury was found to be a predictor of severe complications (p = 0.027).

Conclusion

These findings confirm the clinical efficacy and safety during the implementation phase of POEM for all EMD in Spain.

Introduction

Anti-reflux surgery is a treatment option for patients with gastroesophageal reflux disease (GERD) in which medical therapy has failed, is not tolerated, or is not desired. Conventional surgical interventions have well-known sequelae non grata, such as postoperative dysphagia, possibly more severe in subjects with ineffective esophageal motility (IEM), and hiatal hernia (HH) recurrence, for which a large HH (>3 cm) at baseline is a risk factor.^1-3 These real-world disease features complicate optimal treatment. RefluxStop is an emerging technology that treats GERD without encircling the lower esophagus (avoiding dysphagia) and acts as a mechanical stop preventing re-herniation.⁴ This study aims to present 12-month outcomes of the first 100 patients at our center in Switzerland.

Aims & Methods

A retrospective analysis of the first 100 patients (June 2020-December 2023) that underwent RefluxStop surgery with 12-month follow-up. Clinical outcomes included GERD Health-Related Quality of Life (GERD-HRQL) scores, patient satisfaction, proton pump inhibitor (PPI) use, and postoperative adverse outcomes.

Results

Patient characteristics are shown in Table 1. Median (IQR) total GERD-HRQL score (0-75) decreased from 41 (28.75-50) at baseline to 1 (0-5) at 12 months (p<.001) with patient satisfaction fulfilled in 92% after surgery. Non-reliance on PPIs increased from 5% at baseline to 92% at follow-up. One case of early device penetration (1%) occurred likely from overly tight closure of the invagination pouch, without action required, an outcome attributed to the learning curve of a new procedure. The device is designed to disassociate into its five components and pass uneventfully through the digestive tract. One dislocation (1%) occurred from improper closure of the deployment tool channel (i.e., surgical error). Three cases of re-do hiatus repair (3%) occurred in patients with very large HH (>7 cm) at baseline, where even the large fundic package (~4 cm) with the invaginated device is not sufficient to prevent recurrence in the case of total hiatus repair collapse.

Table 1. Baseline clinical characteristics

Total patients	N=100
Age, mean (SD), years	55.9 (14.9)
Sex, male, n (%)	54 (54%)
Regular PPI use at baseline, n (%)	95 (95%)
HH size, mean (SD), cm	3.8 (1.7)
Proportion with HH >3 cm, n (%)	55 (55%)
Proportion with IEM, n (%)	66 (66%)

HH, hiatal hernia; IEM, ineffective esophageal motility; IQR, interquartile range; PPI, proton pump inhibitor; SD, standard deviation.

Conclusion

RefluxStop showed excellent 12-month results in the treatment of GERD despite the inclusion of difficult-to-treat patients with esophageal dysmotility (66%) and large HH >3 cm in size (55%). Quality of life (i.e., GERD-HRQL) substantially improved by 97.6% from score 41 to 1, with minimal adverse effects. This data contributes to the growing body of evidence supporting the benefits of RefluxStop in the surgical treatment of GERD patients.

References

1. Baison GN, Jackson AS, Wilshire CL, et al. The Impact of Ineffective Esophageal Motility on Patients Undergoing Magnetic Sphincter Augmentation. Ann Surg. 2023;277(4):e793-e800.
2. Hasak S, Brunt LM, Wang D, Gyawali CP. Clinical Characteristics and Outcomes of Patients With Postfundoplication Dysphagia. Clin Gastroenterol Hepatol. 2019;17(10):1982-90.
3. Armijo PR, Pokala B, Misfeldt M, et al. Predictors of Hiatal Hernia Recurrence After Laparoscopic Anti-reflux Surgery with Hiatal Hernia Repair: a Prospective Database Analysis. J Gastrointest Surg. 2019;23(4):696-701.
4. Bonavina L. Individualizing surgical therapy for gastroesophageal reflux disease. Curr Opin Gastroenterol. 2025.

Disclosure

Ioannis Linas and Joerg Zehetner are consultants for Implantica (producer of the RefluxStop device).