Introduction

Capsule endoscopy (CE) is commonly used as the initial exam in situations of suspected mid-gastrointestinal bleeding, after normal upper and lower endoscopy. Although the assessment of the small bowel is the primary focus of CE, detection of upstream or downstream vascular lesions may also be clinically significant. This study aimed to develop and test a Convolutional Neural Network (CNN)-based model for panendoscopic automatic detection of vascular lesions during CE.

Aims & Methods

A multicentric retrospective study was conducted, based on 1022 CE exams. We used a total of 35655 frames, from seven types of CE devices (PillCam Colon 1, PillCam Crohn's, PillCam SB1, PillCam SB3, OMOM HD, Olympus, Mirocam), of which 11091 were considered to have vascular lesions (angiectasia or varices) after triple validation. We divided data into a training and a validation set, the latter being used to evaluate the model’s performance. At time of division, all frames from a given patient were assigned to the same dataset. Our primary outcome measures were sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV) and an area under the precision-recall curve (AUC-PR).

Results

A total of 35655 frames were used, with 11091 including vascular lesions. The validation data set contained 2433 frames with normal mucosa and 933 vascular lesions.
Sensitivity and specificity were 86.4% and 98.3%, respectively. PPV was 95.2%, while the NPV was 95.0%. Overall accuracy was 95.0%. The AUC-PR value was 0.96. The CNN processed 115 frames per second.

Conclusion

The study presents promising results for the application of AI deep learning models in panendoscopic automatic detection of vascular lesions in various gastrointestinal topographies. The methodological details, including the patient-split design and the use of frames from different types of CE devices, suggest that the model could be effectively used in routine clinical practice independently of the device brand. This study highlights the potential of panendoscopic evaluation of all gastrointestinal tract and the need for further research in this area. Overall, the high diagnostic performance of this CNN in multibrand devices addresses an important issue of technological interoperability, allowing it to be replicated in multiple technological settings. However, the study has some limitations, including a small number of frames and a retrospective design, which may introduce selection bias, therefore prospective and multicentric studies are needed before introducing this deep learning model into clinical practice. This is the first proof-of concept AI deep learning model developed for panendoscopic automatic detection of vascular lesions during CE.

References

Triester, S.L., et al., A meta-analysis of the yield of capsule endoscopy compared to other diagnostic modalities in patients with obscure gastrointestinal bleeding. Am J Gastroenterol, 2005. 100(11): p. 2407-18.
Liao, Z., et al., Indications and detection, completion, and retention rates of small-bowel capsule endoscopy: a systematic review. Gastrointest Endosc, 2010. 71(2): p. 280-6.
Rondonotti, E., et al., Small-bowel capsule endoscopy and device-assisted enteroscopy for diagnosis and treatment of small-bowel disorders: European Society of Gastrointestinal Endoscopy (ESGE) Technical Review. Endoscopy, 2018. 50(4): p. 423-446.
Ribeiro, T.F.-U.I.C.H., Colon capsule endoscopy and artificial intelligence: a perfect match por panendoscopy, in Artificial Intelligence in Capsule Endoscopy, C.H. Mascarenhas M, Macedo G, Editor. 2023, Academic Press.
Tsuboi, A., et al., Artificial intelligence using a convolutional neural network for automatic detection of small-bowel angioectasia in capsule endoscopy images. Dig Endosc, 2020. 32(3): p. 382-390.
Saurin, J.C., et al., Diagnostic value of endoscopic capsule in patients with obscure digestive bleeding: blinded comparison with video push-enteroscopy. Endoscopy, 2003. 35(7): p. 576-84.
Mascarenhas, M., et al., Deep learning and colon capsule endoscopy: automatic detection of blood and colonic mucosal lesions using a convolutional neural network. Endosc Int Open, 2022. 10(2): p. E171-e177