Magnetically controlled capsule endoscopy for the evaluation of the stomach. Are we ready for this?

Development and Verification of a Deep Learning Algorithm to Evaluate Small-Bowel Preparation Quality

Capsule endoscopy (CE) quality control requires an objective scoring system to evaluate the preparation of the small bowel (SB). We propose a deep learning algorithm to calculate SB cleansing scores and verify the algorithm’s performance. A 5-point scoring system based on clarity of mucosal visualization was used to develop the deep learning algorithm (400,000 frames; 280,000 for training and 120,000 for testing). External validation was performed using additional CE cases (n = 50), and average cleansing scores (1.0 to 5.0) calculated using the algorithm were compared to clinical grades (A to C) assigned by clinicians. Test results obtained using 120,000 frames exhibited 93% accuracy. The separate CE case exhibited substantial agreement between the deep learning algorithm scores and clinicians’ assessments (Cohen’s kappa: 0.672). In the external validation, the cleansing score decreased with worsening clinical grade (scores of 3.9, 3.2, and 2.5 for grades A, B, and C, respectively, p < 0.001). Receiver operating characteristic curve analysis revealed that a cleansing score cut-off of 2.95 indicated clinically adequate preparation. This algorithm provides an objective and automated cleansing score for evaluating SB preparation for CE. The results of this study will serve as clinical evidence supporting the practical use of deep learning algorithms for evaluating SB preparation quality.

Video capsule endoscopy for upper gastrointestinal hemorrhage in the emergency department: A systematic review and meta-analysis

OBJECTIVE

The assessment of the severity of upper gastrointestinal hemorrhage in emergency department (ED) patients is difficult to assess with commonly available diagnostic tools. Small studies have shown that video capsule endoscopy (VCE) is a promising risk-stratification method and may be better than current clinical decision rules such as the Rockall score and the Glasgow Blatchford score. This review aims to assess the accuracy of VCE to detect active upper gastrointestinal hemorrhage compared to a reference standard.

METHODS

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) methodology was used to perform a review of studies that have measured the diagnostic accuracy of VCE. Studies were included if they measured ED use of VCE for upper GI hemorrhage as compared to a reference standard of an esophagogastroduodenoscopy (EGD). A meta-analysis was performed on select patients using a fixed effects and random-effects model to determine the primary outcome of diagnostic test accuracy.

RESULTS

40 studies were screened for eligibility and five studies representing 193 patients met the inclusion and exclusion criteria. All patients received both a VCE and an EGD. The sensitivity and specificity of VCE were 0.724 and 0.748, respectively. The diagnostic odds ratio was 6.29 (95% CI: 3.23-12.25) and the summary receiver operating characteristic curve was 0.782.

CONCLUSIONS

VCE demonstrated high accuracy for detecting upper GI hemorrhage in this meta-analysis of existing studies. In light of the potential advantages of VCE in the ED, further research is warranted to further establish its role.