AI-assisted capsule endoscopy reading in suspected small bowel bleeding: a multicentre prospective study.

BACKGROUND: Capsule endoscopy reading is time consuming, and readers are required to maintain attention so as not to miss significant findings. Deep convolutional neural networks can recognise relevant findings, possibly exceeding human performances and reducing the reading time of capsule endoscopy. Our primary aim was to assess the non-inferiority of artificial intelligence (AI)-assisted reading versus standard reading for potentially small bowel bleeding lesions (high P2, moderate P1; Saurin classification) at per-patient analysis. The mean reading time in both reading modalities was evaluated among the secondary endpoints. METHODS: Patients aged 18 years or older with suspected small bowel bleeding (with anaemia with or without melena or haematochezia, and negative bidirectional endoscopy) were prospectively enrolled at 14 European centres. Patients underwent small bowel capsule endoscopy with the Navicam SB system (Ankon, China), which is provided with a deep neural network-based AI system (ProScan) for automatic detection of lesions. Initial reading was performed in standard reading mode. Second blinded reading was performed with AI assistance (the AI operated a first-automated reading, and only AI-selected images were assessed by human readers). The primary endpoint was to assess the non-inferiority of AI-assisted reading versus standard reading in the detection (diagnostic yield) of potentially small bowel bleeding P1 and P2 lesions in a per-patient analysis. This study is registered with ClinicalTrials.gov, NCT04821349. FINDINGS: From Feb 17, 2021 to Dec 29, 2021, 137 patients were prospectively enrolled. 133 patients were included in the final analysis (73 [55%] female, mean age 66·5 years [SD 14·4]; 112 [84%] completed capsule endoscopy). At per-patient analysis, the diagnostic yield of P1 and P2 lesions in AI-assisted reading (98 [73·7%] of 133 lesions) was non-inferior (p<0·0001) and superior (p=0·0213) to standard reading (82 [62·4%] of 133; 95% CI 3·6-19·0). Mean small bowel reading time was 33·7 min (SD 22·9) in standard reading and 3·8 min (3·3) in AI-assisted reading (p<0·0001). INTERPRETATION: AI-assisted reading might provide more accurate and faster detection of clinically relevant small bowel bleeding lesions than standard reading. FUNDING: ANKON Technologies, China and AnX Robotica, USA provided the NaviCam SB system.

Endoscopic Gastrointestinal Placement of Capsule Endoscopy to Investigate the Small Bowel: a Multicentric European Retrospective Series of 630 Procedures in Adult Patients.

BACKGROUND AND AIMS: Small bowel (SB) capsule endoscopy (CE) is a first line procedure for exploring the SB. Endoscopic GastroIntestinal PlacemenT (EGIPT) of SB CE is sometimes necessary. While the experience of EGIPT is large in pediatric populations, we aimed to describe the safety, efficacy and outcomes of EGIPT of SB CE in adult patients. METHODS: The international CApsule endoscopy REsearch (iCARE) group set up a retrospective multicenter study. Patients over 18 year-old who underwent EGIPT of SB CE before May 2022 were included. Data were collected from medical records and capsule recordings. The primary endpoint was the technical success rate of the EGIPT procedures. RESULTS: 630 patients were included (mean age 62.5 years old, 55.9% female) from 39,565 patients (1.6%) issued from 29 centers. EGIPT technical success was achieved in 610 procedures (96.8%). Anesthesia (moderate/deep sedation or general anesthesia) and centers with intermediate or high procedure loads were independent factors of technical success. Severe adverse events occurred in three (0.5%) patients. When technically successful, EGIPT was associated with a high SB CE completion rate (84.4%) and with a substantial diagnostic yield (61.1%). Completion rate was significantly higher when the capsule was delivered in the SB compared to when delivered in the stomach. CONCLUSION: EGIPT of SB CE is highly feasible, safe and comes with high completion rate and diagnostic yield. When indicated, it should rather be performed under anesthesia and the capsule should be delivered in the duodenum rather than in the stomach, for better SB examination outcomes.

First prospective European study for the feasibility and safety of magnetically controlled capsule endoscopy in gastric mucosal abnormalities.

BACKGROUND: While capsule endoscopy (CE) is the gold standard diagnostic method of detecting small bowel (SB) diseases and disorders, a novel magnetically controlled capsule endoscopy (MCCE) system provides non-invasive evaluation of the gastric mucosal surface, which can be performed without sedation or discomfort. During standard SBCE, passive movement of the CE may cause areas of the complex anatomy of the gastric mucosa to remain unexplored, whereas the precision of MCCE capsule movements inside the stomach promises better visualization of the entire mucosa. AIM: To evaluate the Ankon MCCE system's feasibility, safety, and diagnostic yield in patients with gastric or SB disorders. METHODS: Of outpatients who were referred for SBCE, 284 (male/female: 149/135) were prospectively enrolled and evaluated by MCCE. The stomach was examined in the supine, left, and right lateral decubitus positions without sedation. Next, all patients underwent a complete SBCE study protocol. The gastric mucosa was explored with the Ankon MCCE system with active magnetic control of the capsule endoscope in the stomach, applying three standardized pre-programmed computerized algorithms in combination with manual control of the magnetic movements. RESULTS: The urea breath test revealed Helicobacter pylori positivity in 32.7% of patients. The mean gastric and SB transit times with MCCE were 0 h 47 min 40 s and 3 h 46 min 22 s, respectively. The average total time of upper gastrointestinal MCCE examination was 5 h 48 min 35 s. Active magnetic movement of the Ankon capsule through the pylorus was successful in 41.9% of patients. Overall diagnostic yield for detecting abnormalities in the stomach and SB was 81.9% (68.6% minor; 13.3% major pathologies); 25.8% of abnormalities were in the SB; 74.2% were in the stomach. The diagnostic yield for stomach/SB was 55.9%/12.7% for minor and 4.9%/8.4% for major pathologies. CONCLUSION: MCCE is a feasible, safe diagnostic method for evaluating gastric mucosal lesions and is a promising non-invasive screening tool to decrease morbidity and mortality in upper gastro-intestinal diseases.

Improved adenoma detection with linked color imaging technology compared to white-light colonoscopy.

OBJECTIVES: Linked color imaging (LCI) is a new endoscopic technology that may increase colorectal adenoma detection rate (ADR) and polyp detection rate (PDR) by virtual chromoendoscopy. Aim of the present study was to evaluate the effectiveness of LCI in ADR and PDR compared to the HD white-light colonoscopy (WLC) technique. MATERIALS AND METHODS: Between October 2016 and June 2018, we enrolled consecutive outpatients prospectively. Eligible patients allocated randomly to undergo HD WLC or LCI colonoscopy technique during instrument withdrawal. Each colonoscopy was performed in a single center by the same three expert endoscopists (with expertise more than 5000 colonoscopies). RESULTS: A total of 1278 patients underwent colonoscopy in the study period. ADR and PDR were significantly higher in the LCI group compared to the WLC group (34.4% vs. 26.8%; p = .007; and 53.3% vs 46.4%; p = .023, respectively). Similarly, the mean number of adenomas per patient (MAP) was significantly higher with the LCI than WLC (0.64 vs 0.44, respectively; p = .002). The mean age of patients at the time of colonoscopy was 51.95 years (SD = 13.861) in the LCI group and 51.96 years (SD = 14.028) in the WLC group. No significant differences observed in patient demographic characteristics (there was no difference in gender and age distribution, p = .986), quality of colonoscopy preparation and withdrawal times (WT) between the two groups (WLC and the LCI groups, 493.9 (SD: 143.5) and 514.0 (SD: 139.5) sec, respectively). CONCLUSIONS: According to our results, LCI virtual chromoendoscopic technology was superior compared to conventional HD WLC in detecting colorectal polyps and adenomas.