Introduction

The diagnosis of small bowel motility disorders currently relies on identifying abnormal contractile patterns through intestinal manometry, but the technique is restricted to referral centers. Conventional CT scanning, frequently performed in the routine evaluation of these patients, provides images of intestinal content, which might serve as a marker of intestinal dysfunction. We hypothesized that pooling of intestinal content, quantified through computerized analysis of abdominal CT images, could serve as an indirect marker of small intestinal motor dysfunction.

Aims & Methods

Consecutive patients with either a) chronic intestinal pseudo-obstruction (CIPO) or b) suspected small intestinal motor dysfunction, evaluated at our Motility Unit using high-resolution jejunal manometry, were prospectively included. Manometric tracings were categorized as
a) normal,
b) uncoordinated contractile activity with preserved amplitude, or
c) hypomotility (markedly diminished amplitude of contractile activity).
All patients also underwent an abdominal CT scan within three months of the manometric study, as part of the clinical evaluation of the motility disorder. CT images were analyzed using a validated software developed in our laboratory, which applies image processing techniques and deep-learning tools to identify and quantify intestinal content volumes in the small bowel and colon of:
a) gas,
b) chyme-type content (non-gaseous, homogeneous texture), and
c) fecal-type content (non-gaseous, heterogeneous texture).

Results

Of the 66 patients included, 9 had CIPO and 57 suspected dysmotility. Manometry was abnormal in all patients with CIPO and in 22 patients (39%) without CIPO. Total intestinal content was markedly increased in patients with CIPO as compared with patients with normal motility (3034±826 mL vs 398±43 mL, respectively; p<0.001), primarily due to an increase in gas (1086±325 mL vs 22±4; p<0.001) and chyme-type content in the small bowel (1282±544 mL vs 81±11; p<0.001).
Among patients without CIPO, the volume of small bowel content was increased in patients with dysmotility compared to those with normal motility, mainly of chyme-type content (191±39 vs 81±11 mL in normal motility; p=0.006). Colonic volumes of gas and fecal-type content were also higher in patients with dysmotility than in those with normal motility (gas 224±40 vs 115±18 mL, p=0.027; fecal-type 244±59 vs 115±19 mL, p=0.036). Dysmotility patients with hypomotility patterns had overall higher intestinal content volumes than patients with uncoordinated contractile patterns (Table 1).

Table 1. Intestinal content volumes in non-CIPO patients, classified by manometric patterns

Conclusion

Intestinal content, measured on conventional CT scans by a specific analysis program, provides a useful indication of intestinal dysmotility.

Introduction

While anatomical landmark identification and adequate inspection time are recognized as crucial quality indicators in upper endoscopy, they are difficult to document consistently and accurately based solely on endoscopist reports.

Aims & Methods

To address this issue at Assuta Healthcare, a multicenter organization with 20 endoscopy units and over 200 endoscopists, we implemented a video-based AI platform designed to measure total upper endoscopy time (EGD-T), assess photo documentation of landmarks, and streamline the photo labeling process. In the current study, we evaluated the platform and employed it to examine endoscopist-reported quality measures.
For all upper endoscopies performed at Assuta Healthcare in 2024 on patients 45 years and older, the Theator AI-based Surgical Intelligence platform automatically documented total procedure time as well as photo snapshots captured and labeled by endoscopists for seven landmarks: 1. esophagus, 2. gastroesophageal junction, 3. cardia or fundus, 4. body, 5. angulus, 6. antrum or pylorus, and 7. duodenum (second part). In a subset of procedures, the platform facilitated documentation by suggesting labels for captured snapshots. After validating platform accuracy, we examined platform-reported procedure times. We then assessed landmark photo documentation rates in procedures with and without platform labeling suggestions.

Results

The study included 21,989 upper endoscopies (60.7% female, median age of 62.2 years), of which 51.7% were conducted with platform labeling suggestions and the remaining 48.3% were conducted without. Validated platform accuracy was 98%. Median total procedure time measured by the platform was 3.18 minutes (IQR 2.18-4.25). When they did not receive suggestions, endoscopists did not label any landmarks in 25% of procedures. At 12%, this percentage was significantly lower (p < .001) when the platform did provided suggestions. This difference was also reflected in the median number of landmarks labeled, which was 2 (IQR 0-3) in procedures without labeling suggestions and 3 (IQR 1-4) in procedures with labeling suggestions (p < .001).

Conclusion

Implementing an AI tool to automatically assess procedure time in upper endoscopy provided accurate measurements, revealing durations considerably lower than recommended. It further revealed extremely low labeling rates, which increased when automated suggestions were provided. The results highlight critical gaps in the reporting of quality measures for upper endoscopy and indicate that AI-based documentation can be instrumental to quality assurance by identifying and addressing such gaps.

Disclosure

None

Introduction

Artificial Intelligence (AI) assisted reading in Small Bowel Capsule Endoscopy (SBCE) has recently been shown to achieve comparable and potentially superior accuracy compared to standard clinician reading. In Colon Capsule Endoscopy (CCE), AI algorithms have also demonstrated promising results. However, the extent of AI-assisted reading's advantage remains unclear, particularly regarding its performance across different polyp sizes, morphologies, locations, and other factors. A deeper understanding of these variables is crucial for optimising AI performance and facilitating its clinical integration.

Aims & Methods

This CESCAIL sub-analysis evaluates the per-polyp diagnostic accuracy of AI-assisted versus standard clinician assessment (readings) and identifies key factors influencing AI-assisted accuracy using AiSPEED^TM, the Computer-Aided Detection (CADe) system.
The CESCAIL study evaluated the diagnostic accuracy of AI-assisted CCE readings using AiSPEED™ compared to the standard readings. Standard readings were manually performed by clinicians, while AI-assisted readings incorporated an AI pre-review followed by clinician validation. A total of 1,803 polyps from 673 patients were analysed at the per-polyp level, focusing on sensitivity and positive predictive value. Factors influencing the improved accuracy of AI-assisted readings over standard clinician readings were also assessed, including polyp size, morphology, location, patient demographics (age and sex), bowel preparation quality, capsule excretion rates, comorbidities, medications, reading time, and video duration.

Results

AI-assisted reading demonstrated significantly higher sensitivity with clear superiority for smaller polyps (<10 mm) compared to larger ones (≥10 mm) (OR 2.27 vs 0.88, p<0.001) (see Table 1). While there was no observed difference in diagnostic accuracy between AI-assisted and standard clinician assessment for polyps ≥10 mm, non-inferiority was established. AI accuracy remained consistent between 6–9 mm and ≤5 mm polyps (p=0.64). The most notable improvement was observed with hyperplastic polyps (HPs) (OR 5.4, p<0.001). No significant differences were identified for pedunculated, sub-pedunculated, Lateral Spreading Tumours, or Sessile Serrated Lesions. Furthermore, AI-assisted readings were significantly more accurate for left-sided polyps than right-sided ones (OR 2.36 vs 1.66, p<0.0001), although AI-assisted reads outperformed standard reads in both locations.

Conclusion

This study highlights the performance and consistency of AI-assisted reading, particularly for detecting smaller adenomas and HPs, with notably improved accuracy in the left colon. Next-generation AI should focus on distinguishing significant from diminutive polyps and enhancing polyp characterisation, especially for right-sided lesions.

References

1. Spada C, Piccirelli S, Hassan C, Ferrari C, Toth E, Gonzalez-Suarez B, et al. AI-assisted capsule endoscopy reading in suspected small bowel bleeding: a multicentre prospective study. Lancet Digit Health. 2024;6(5):e345-e53.

Disclosure

Hagen Wenzek serves as the CEO, Elizabeth White is the Project Manager, and Pablo Laiz is the AI software engineer at Corporate Health International, which owns the intellectual properties of AiSPEED^TM.

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

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.

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