Artificial intelligence using a convolutional neural network for automatic detection of small-bowel angioectasia in capsule endoscopy images
Akiyoshi Tsuboi
Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan
Search for more papers by this authorCorresponding Author
Shiro Oka
Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan
Corresponding: Shiro Oka, Department of Gastroenterology and Metabolism, Hiroshima University Hospital, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8551, Japan. Email: [email protected]Search for more papers by this authorHiroaki Saito
Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan
Search for more papers by this authorTomoki Matsuda
Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan
Search for more papers by this authorMitsuhiro Fujishiro
Department of Gastroenterology & Hepatology, Nagoya University Graduate School of Medicine, Aichi, Japan
Search for more papers by this authorSoichiro Ishihara
Department of Surgical Oncology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
Tada Tomohiro Institute of Gastroenterology and Proctology, Saitama, Japan
Search for more papers by this authorMasato Nakahori
Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan
Search for more papers by this authorShinji Tanaka
Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan
Search for more papers by this authorTomohiro Tada
AI Medical Service Inc., Tokyo, Japan
Department of Surgical Oncology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
Tada Tomohiro Institute of Gastroenterology and Proctology, Saitama, Japan
Search for more papers by this authorAkiyoshi Tsuboi
Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan
Search for more papers by this authorCorresponding Author
Shiro Oka
Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan
Corresponding: Shiro Oka, Department of Gastroenterology and Metabolism, Hiroshima University Hospital, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8551, Japan. Email: [email protected]Search for more papers by this authorHiroaki Saito
Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan
Search for more papers by this authorTomoki Matsuda
Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan
Search for more papers by this authorMitsuhiro Fujishiro
Department of Gastroenterology & Hepatology, Nagoya University Graduate School of Medicine, Aichi, Japan
Search for more papers by this authorSoichiro Ishihara
Department of Surgical Oncology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
Tada Tomohiro Institute of Gastroenterology and Proctology, Saitama, Japan
Search for more papers by this authorMasato Nakahori
Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan
Search for more papers by this authorShinji Tanaka
Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan
Search for more papers by this authorTomohiro Tada
AI Medical Service Inc., Tokyo, Japan
Department of Surgical Oncology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
Tada Tomohiro Institute of Gastroenterology and Proctology, Saitama, Japan
Search for more papers by this authorAbstract
Background and Aim
Although small-bowel angioectasia is reported as the most common cause of bleeding in patients and frequently diagnosed by capsule endoscopy (CE) in patients with obscure gastrointestinal bleeding, a computer-aided detection method has not been established. We developed an artificial intelligence system with deep learning that can automatically detect small-bowel angioectasia in CE images.
Methods
We trained a deep convolutional neural network (CNN) system based on Single Shot Multibox Detector using 2237 CE images of angioectasia. We assessed its diagnostic accuracy by calculating the area under the receiver operating characteristic curve (ROC-AUC), sensitivity, specificity, positive predictive value, and negative predictive value using an independent test set of 10 488 small-bowel images, including 488 images of small-bowel angioectasia.
Results
The AUC to detect angioectasia was 0.998. Sensitivity, specificity, positive predictive value, and negative predictive value of CNN were 98.8%, 98.4%, 75.4%, and 99.9%, respectively, at a cut-off value of 0.36 for the probability score.
Conclusions
We developed and validated a new system based on CNN to automatically detect angioectasia in CE images. This may be well applicable to daily clinical practice to reduce the burden of physicians as well as to reduce oversight.
Conflicts of Interest
Authors declare no conflicts of interests for this article.
Supporting Information
| Filename | Description |
|---|---|
| den13507-sup-0001-VideoS1.mp4MPEG-4 video, 14.4 MB | Video S1 Video demonstration of real-time detection for small-bowel angioectasia by convolutional neural network (CNN). Representative video detected multiple small-bowel angioectasias by CNN using real video files (yellow box, region identified as an angioectasia by CNN; number, the probability score of CNN reading). |
Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
References
- 1Iddan G, Meron G, Glukhovsky A et al. Wireless capsule endoscopy. Nature 2000; 405: 417.
- 2Meron GD. The development of the swallowable video capsule (M2A). Gastrointest. Endosc. 2000; 52: 817–9.
- 3Lewis BS, Swain P. Capsule endoscopy in the evaluation of patients with suspected small intestinal bleeding: results of a pilot study. Gastrointest. Endosc. 2002; 56: 349–53.
- 4Pennazio M, Santucci R, Rondonotti E et al. Outcome of patients with obscure gastrointestinal bleeding after capsule endoscopy: report of 100 consecutive cases. Gastroenterology 2004; 126: 643–53.
- 5Apostolopoulos P, Liatsos C, Gralnek IM et al. The role of wireless capsule endoscopy in investigating unexplained iron deficiency anemia after negative endoscopic evaluation of the upper and lower gastrointestinal tract. Endoscopy 2006; 38: 1127–32.
- 6Apostolopoulos P, Liatsos C, Gralnek IM et al. Evaluation of capsule endoscopy in active, mild-to-moderate, overt, obscure GI bleeding. Gastrointest. Endosc. 2007; 66: 1174–81.
- 7Urbain D, De Looze D, Demedts I et al. Video capsule endoscopy in small-bowel malignancy: a multicenter Belgian study. Endoscopy 2006; 38: 408–11.
- 8Bailey AA, Debinski HS, Appleyard MN et al. Diagnosis and outcome of small bowel tumors found by capsule endoscopy: a three-center Australian experience. Am. J. Gastroenterol. 2006; 101: 2237–43.
- 9Hartmann D, Schmidt H, Bolz G et al. A prospective two-center study comparing wireless capsule endoscopy with intraoperative enteroscopy in patients with obscure GI bleeding. Gastrointest. Endosc. 2005; 61: 826–32.
- 10Imagawa H, Oka S, Tanaka S et al. Improved detectability of small-bowel lesions via capsule endoscopy with computed virtual chromoendoscopy: a pilot study. Scand. J. Gastroenterol. 2011; 46: 1133–7.
- 11Watari I, Oka S, Tanaka S et al. Is occult obscure gastrointestinal bleeding a definite indication for capsule endoscopy? A retrospective analysis of diagnostic yield in patients with occult versus overt bleeding. Gastroenterol. Res. Pract. 2013; 2013: 915463.
- 12McAlindon ME, Ching HL, Yung D et al. Capsule endoscopy of the small bowel. Ann. Transl. Med. 2016; 4: 369.
- 13Schmit A, Gay F, Adler M et al. Diagnostic efficacy of push-enteroscopy and long-term follow-up of patients with small bowel angiodysplasias. Dig. Dis. Sci. 1996; 41: 2348–52.
- 14Yano T, Yamamoto H, Sunada K et al. Endoscopic classification of vascular lesions of the small intestine (with videos). Gastrointest. Endosc. 2008; 67: 169–72.
- 15Gulshan V, Peng L, Coram M et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA 2016; 316: 2402–10.
- 16Esteva A, Kuprel B, Novoa RA et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature 2017; 542: 115–8.
- 17Ehteshami Bejnordi B, Veta M, Johannes van Diest P et al. Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. JAMA 2017; 318: 2199–210.
- 18Shichijo S, Nomura S, Aoyama K et al. Application of convolutional neural networks in the diagnosis of Helicobacter pylori infection based on endoscopic images. EBioMedicine 2017; 25: 106–11.
- 19Hirasawa T, Aoyama K, Tanimoto T et al. Application of artificial intelligence using a convolutional neural network for detecting gastric cancer in endoscopic images. Gastric Cancer 2018; 21: 653–60.
- 20LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015; 521: 436–44.
- 21Shin HC, Roth HR, Gao M et al. Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning. IEEE Trans. Med. Imaging 2016; 35: 1285–98.
- 22Xiao J, Meng MQ. A deep convolutional neural network for bleeding detection in wireless capsule endoscopy images. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2016; 2016: 639–42.
- 23Zhou T, Han G, Li BN et al. Quantitative analysis of patients with celiac disease by video capsule endoscopy: a deep learning method. Comput. Biol. Med. 2017; 85: 1–6.
- 24Noya F, Alvarez-Gonzalez MA, Benitez R. Automated angiodysplasia detection from wireless capsule endoscopy. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2017; 2017: 3158–61.
- 25Leenhardt R, Vasseur P, Li C et al. A neural network algorithm for detection of GI angiectasia during small-bowel capsule endoscopy. Gastrointest. Endosc. 2019; 89: 189–94.
- 26Mylonaki M, Fritscher-Ravens A, Swain P. Wireless capsule endoscopy: a comparison with push enteroscopy in patients with gastroscopy and colonoscopy negative gastrointestinal bleeding. Gut 2003; 52: 1122–6.
- 27Delvaux M, Fassler I, Gay G. Clinical usefulness of the endoscopic video capsule as the initial intestinal investigation in patients with obscure digestive bleeding: validation of a diagnostic strategy based on the patient outcome after 12 months. Endoscopy 2004; 36: 1067–73.
- 28Li F, Leighton JA, Sharma VK. Capsule endoscopy in the evaluation of obscure gastrointestinal bleeding: a comprehensive review. Gastroenterol. Hepatol. (NY) 2007; 3: 777–85.
- 29Imagawa H, Oka S, Tanaka S et al. Improved visibility of lesions of the small intestine via capsule endoscopy with computed virtual chromoendoscopy. Gastrointest. Endosc. 2011; 73: 299–306.
- 30Minami-Kobayashi Y, Yamada A, Watabe H et al. Efficacy of repeat review with flexible spectral imaging color enhancement in patients with no findings by capsule endoscopy. Saudi J. Gastroenterol. 2016; 22: 385–90.
- 31Yung DE, Boal Carvalho P, Giannakou A et al. Clinical validity of flexible spectral imaging color enhancement (FICE) in small-bowel capsule endoscopy: a systematic review and meta-analysis. Endoscopy 2017; 49: 258–69.
- 32Aoki T, Yamada A, Aoyama K et al. Automatic detection of erosions and ulcerations in wireless capsule endoscopy images based on a deep convolutional neural network. Gastrointest. Endosc. 2019; 89: 357–63.e2.
- 33Karagiannis S, Goulas S, Kosmadakis G et al. Wireless capsule endoscopy in the investigation of patients with chronic renal failure and obscure gastrointestinal bleeding (preliminary data). World J. Gastroenterol. 2006; 12: 5182–5.
- 34Bhutani MS, Gupta SC, Markert RJ et al. A prospective controlled evaluation of endoscopic detection of angiodysplasia and its association with aortic valve disease. Gastrointest. Endosc. 1995; 42: 398–402.
- 35Rosborough TK, Swaim WR. Acquired von Willebrand's disease, platelet-release defect and angiodysplasia. Am. J. Med. 1978; 65: 96–100.
- 36Igawa A, Oka S, Tanaka S et al. Major predictors and management of small-bowel angioectasia. BMC Gastroenterol. 2015; 15: 108.
- 37Yamada A, Niikura R, Kobayashi Y et al. Risk factors for small bowel angioectasia: the impact of visceral fat accumulation. World J. Gastroenterol. 2015; 21: 7242–7.
- 38Pan G, Yan G, Song X et al. BP neural network classification for bleeding detection in wireless capsule endoscopy. J. Med. Eng. Technol. 2009; 33: 575–81.
- 39Li B, Meng MQ. Computer-aided detection of bleeding regions for capsule endoscopy images. IEEE Trans. Biomed. Eng. 2009; 56: 1032–9.
- 40Fu Y, Zhang W, Mandal M et al. Computer-aided bleeding detection in WCE video. IEEE J. Biomed. Health Inform. 2014; 18: 636–42.
Citing Literature
