Artificial intelligence for the real-time classification of intrapapillary capillary loop patterns in the endoscopic diagnosis of early oesophageal squamous cell carcinoma: A proof-of-concept study.

Background: Intrapapillary capillary loops (IPCLs) represent an endoscopically visible feature of early squamous cell neoplasia (ESCN) which correlate with invasion depth - an important factor in the success of curative endoscopic therapy. IPCLs visualised on magnification endoscopy with Narrow Band Imaging (ME-NBI) can be used to train convolutional neural networks (CNNs) to detect the presence and classify staging of ESCN lesions. Methods: A total of 7046 sequential high-definition ME-NBI images from 17 patients (10 ESCN, 7 normal) were used to train a CNN. IPCL patterns were classified by three expert endoscopists according to the Japanese Endoscopic Society classification. Normal IPCLs were defined as type A, abnormal as B1-3. Matched histology was obtained for all imaged areas. Results: This CNN differentiates abnormal from normal IPCL patterns with 93.7% accuracy (86.2% to 98.3%) and sensitivity and specificity for classifying abnormal IPCL patterns of 89.3% (78.1% to 100%) and 98% (92% to 99.7%), respectively. Our CNN operates in real time with diagnostic prediction times between 26.17 ms and 37.48 ms. Conclusion: Our novel and proof-of-concept application of computer-aided endoscopic diagnosis shows that a CNN can accurately classify IPCL patterns as normal or abnormal. This system could be used as an in vivo, real-time clinical decision support tool for endoscopists assessing and directing local therapy of ESCN.

Minichromosomal Maintenance Component Complex 5 (MCM5) as a Marker of Barrett's Esophagus-Related Neoplasia: A Feasibility Study.

BACKGROUND: The endoscopic detection of esophageal cancer is suboptimal in both patients referred with dyspeptic symptoms and those enrolled in Barrett's surveillance programs. MCM5 expression in cells collected from gastric fluid may be correlated with the presence of dysplasia or adenocarcinoma. Analysis of this biomarker may improve the detection of cancer. METHODS: Sixty-one patients were enrolled at a single UK referral center. From each patient, 5-10 ml of gastric fluid was aspirated endoscopically. Patients were categorized according to their histology, normal, non-dysplastic Barrett's (NDBE), high-grade dysplastic Barrett's (HGD), and esophageal adenocarcinoma (EAC). All histology was confirmed by Seattle protocol biopsies or endoscopic mucosal resection. Samples were centrifuged, and the cell pellet was lysed. MCM5 expression levels were quantified using a proprietary immunoassay. The mean MCM5 expression was compared between groups by Kruskal-Wallis test. ROC curves were also used to assess diagnostic utility. RESULTS: The mean expression of MCM5 increases as patients progress from a normal esophagus to NDBE, HGD, and EAC (14.4; 49.8; 112.3; and 154.1, respectively). There was a significant difference in the MCM5 expression of patients with a normal esophagus compared to those with EAC (p = 0.04). There was a trend toward higher MCM5 expression in patients with EAC compared to those with NDBE (p = 0.34). MCM5 expression was a fair discriminator (AUC 0.70 [95% CI 0.57-0.83]) between patients without neoplasia (normal and NDBE) and those with early neoplasia (HGD and EAC). CONCLUSION: MCM5 expression in gastric fluid samples can differentiate patients with a histologically normal esophagus compared to those with early adenocarcinoma. Larger, powered studies are needed to assess whether it can be used to differentiate those with HGD from NDBE.