Artificial intelligence‐based detection of pharyngeal cancer using convolutional neural networks. Issue 7 (1st April 2020)

• Record Type: Journal Article
• Authors: Tamashiro, Atsuko; Yoshio, Toshiyuki; Ishiyama, Akiyoshi; Tsuchida, Tomohiro; Hijikata, Kazunori; Yoshimizu, Shoichi; Horiuchi, Yusuke; Hirasawa, Toshiaki; Seto, Akira; Sasaki, Toru; Fujisaki, Junko; Tada, Tomohiro
• Is Part Of: Digestive endoscopy. Volume 32:Issue 7(2020)
• Page Start: 1057
• Page End: 1065
• Publication Date: 2020-04-01
• Subjects: artificial intelligence -- convolutional neural network -- deep learning -- pharyngeal cancer; Digestive organs -- Diseases -- Periodicals; Digestive organs -- Diseases -- Diagnosis -- Periodicals; Endoscopy -- Periodicals; Digestive System Diseases -- diagnosis -- Periodicals; Digestive System Diseases -- therapy -- Periodicals; Endoscopy -- Periodicals; 616.3
• DOI: 10.1111/den.13653
• Languages: English
• ISSNs: 0915-5635

Abstract

Abstract : Objectives: The prognosis for pharyngeal cancer is relatively poor. It is usually diagnosed in an advanced stage. Although the recent development of narrow‐band imaging (NBI) and increased awareness among endoscopists have enabled detection of superficial pharyngeal cancer, these techniques are still not prevalent worldwide. Nevertheless, artificial intelligence (AI)‐based deep learning has led to significant advancements in various medical fields. Here, we demonstrate the diagnostic ability of AI‐based detection of pharyngeal cancer from endoscopic images in esophagogastroduodenoscopy. Methods: We retrospectively collected 5403 training images of pharyngeal cancer from 202 superficial cancers and 45 advanced cancers from the Cancer Institute Hospital, Tokyo, Japan. Using these images, we developed an AI‐based diagnostic system with convolutional neural networks. We prepared 1912 validation images from 35 patients with 40 pharyngeal cancers and 40 patients without pharyngeal cancer to evaluate our system. Results: Our AI‐based diagnostic system correctly detected all pharyngeal cancer lesions (40/40) in the patients with cancer, including three small lesions smaller than 10 mm. For each image, the AI‐based system correctly detected pharyngeal cancers in images obtained via NBI with a sensitivity of 85.6%, much higher sensitivity than that for images obtained via white light imaging (70.1%). The novel diagnostic system took only 28 s to analyze 1912 validation images. Conclusions: The novel AI‐based diagnostic system detected pharyngeal cancer with high sensitivity. It could facilitate early detection, thereby leading to better prognosis and quality of life for patients with pharyngeal cancers in the near future.