Multicentre prospective study on the diagnostic performance of MiroCam MC2000 double tip small bowel capsule AI scan, first real-world study on all indications

Aims Previous studies have shown the superiority of using double-tip capsule endoscope (bi-directional cameras) to increase diagnostic yield in the small bowel, however, this does increase reading times. We aimed to assess the diagnostic performance (sensitivity and specificity) of the MiroCam MC2000 double tip AI scan in detecting diagnostic, in addition to positive findings in the small bowel for all patients referred for small bowel capsule endoscopy. The secondary outcome was to compare MiroCam AI scan reading times with conventional reading [1] [2] [3] [4] [5] [6] [7] [8] [9].

Methods Consecutive patients undergoing small bowel capsule endoscopy for all approved indications at 2 large UK centers from April 2023 to December 2023 were recruited. Capsule endoscopy videos were read by 6 experienced readers from the UK, Italy, Denmark, and South Korea in standard mode. Subsequently, the videos were de-identified, and the same subsets were randomly reallocated to the same readers to be read with MiroCam AI scan. A panel of two expert readers re-read all videos in standard mode and this was used as the reference standard, in addition, they checked for concordance and adjudicated on the diagnostic findings within the context of the full clinical history. Patients who had inadequate bowel preparation were excluded

Results A total of 255 patients were enrolled, and 242 patients were included in the final analysis, mean age of 50.1 years. (SD 18.3) with 53% female (n=128). The most common indication for capsule endoscopy was suspected Crohn's disease 48.8% (n=118) followed by unexplained iron deficiency anaemia 23.6% (n=57), suspected small bowel bleeding 18.6% (n=45), and other 9% (n=22). The sensitivity and specificity of MiroCam AI scan reading for diagnostic findings were 95.3% (CI 90.1%-98.3%) and 96.5% (91.3%-99.0%) respectively. The sensitivity and specificity of the standard reading mode for diagnostic findings were 96.5% (CI 91.2%-99.0%) and 85.3% (78.0%-90.9%) respectively. The MiroCam AI scan had a higher rate of positive findings per patient when compared to standard reading mode 86.0% vs 82.2% (p<0.04). There was a significant reduction in the mean reading time between standard reading mode 38.1 min (SD 20.96) and AI-assisted reading mode 18.26 min (SD 10.79) p<0.01

Conclusions The first double-tip small bowel capsule endoscope with AI, MiroCam MC2000 AI scan, shows high diagnostic accuracy comparable to that of expert readers. It significantly decreases the reading time required for review. This study provides evidence to support the role of AI-assisted reading in the real-world setting covering all approved indications

Conflicts of Interest

Authors do not have any conflict of interest to disclose.

• References

• 1 Messmann H, Bisschops R, Antonelli G. et al. Expected value of artificial intelligence in gastrointestinal endoscopy: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy. 2022; 54: 1211-1231
• 2 Yung DE, Robertson AR, Davie M, Sidhu R, McAlindon M, Rahman I, Patel P. et al. Double-headed small-bowel capsule endoscopy: Real-world experience from a multi-centre British study. Dig Liver Dis 2021; 53 (4): 461-466 Epub 2021 Feb 8. PMID: 33574013
• 3 Hawkes E, Patel P, Gwiggner M. et al PWE-029 Are 2 heads better than 1: randomised comparison of MiroCam single-tip v double-tip capsule endoscope. Gut 2019; 68: A165
• 4 Spada C, Piccirelli S, Hassan C. et al AI-assisted capsule endoscopy reading in suspected small bowel bleeding: a multicentre prospective study. Lancet Digit Health 2024; 6 (5): e345-e353
• 5 Cortegoso Valdivia P, Skonieczna-Żydecka K, Elosua A. et al. Indications, detection, completion and retention rates of capsule endoscopy in two decades of use: a systematic review and meta-analysis. Diagnostics (Basel) 2022; 12: 1105
• 6 Ding Z, Shi H, Zhang H. et al. Artificial intelligence-based diagnosis of abnormalities in small-bowel capsule endoscopy. Endoscopy. 2023; 55: 44-51
• 7 Xie X, Xiao YF, Zhao XY. et al. Development and validation of an artificial intelligence model for small bowel capsule endoscopy video review. JAMA Netw Open 2022; 5: e2221992
• 8 Leenhardt R, Li C, Koulaouzidis A. et al. Nomenclature and semantic description of vascular lesions in small bowel capsule endoscopy: an international Delphi consensus statement. Endosc Int Open 2019; 7 (3): E372-E379
• 9 Leenhardt R, Buisson A, Bourreille A. et al. Nomenclature and semantic descriptions of ulcerative and inflammatory lesions seen in Crohn’s disease in small bowel capsule endoscopy: an international Delphi consensus statement. United European Gastroenterol J 2020; 8 (1): 99-107

Publication History

Article published online:
27 March 2025