Search

Background and aims Endoscopic diagnosis of early esophageal squamous cell cancer (ESCC) is complicated and dependent on operators' experience. This study aimed to develop an artificial intelligence (AI) model for automatic diagnosis of early ESCC. Methods Non-magnifying and magnifying endoscopic images of normal/noncancerous lesions, early ESCC, and advanced esophageal cancer (AEC) were retrospectively obtained from Qilu Hospital of Shandong University. A total of 10,988 images from 5075 cases were chosen for training and validation. Another 2309 images from 1055 cases were collected for testing. One hundred and four real-time videos were also collected to evaluate the diagnostic performance of the AI model. The diagnostic performance of the AI model was compared with endoscopists by magnifying images and the assistant efficiency of the AI model for novices was evaluated. Results The AI diagnosis for non-magnifying images showed a per-patient accuracy, sensitivity, and specificity of 99.5%, 100%, 99.5% for white light imaging, and 97.0%, 97.2%, 96.4% for optical enhancement/iodine straining images. Regarding diagnosis for magnifying images, the per-patient accuracy, sensitivity, and specificity were 88.1%, 90.9%, and 85.0%. The diagnostic accuracy of the AI model was similar to experts (84.5%, P = 0.205) and superior to novices (68.5%, P = 0.005). The diagnostic performance of novices was significantly improved by AI assistance. When it comes to the diagnosis for real-time videos, the AI model showed acceptable performance as well. Conclusions The AI model could accurately recognize early ESCC among noncancerous mucosa and AEC. It could be a potential assistant for endoscopists, especially for novices.

Regional intestinal immune surveillance remains obscure. In this study, we integrated single-cell RNA sequencing and spatial transcriptomics to create a regional atlas of fetal and adult intestines, consisting of 59 cell subsets, of which eight new subsets and ILCs transition states were identified. Results revealed that microenvironment determines in-situ cell differentiation and shapes the regional molecular characteristics, allowing different intestinal segments with diverse functions. We characterized the regional expression of mucins, immunoglobulins, and antimicrobial peptides (AMPs) and their shift during development and in inflammatory bowel disease. Notably, α-defensins expressed most abundantly in small intestinal LGR5+ stem cells, rather than in Paneth cells, and down-regulated as cell maturing. Common upstream transcription factors controlled the AMPs expression, illuminating the concurrent change of AMPs during epithelial differentiation, and the spatial co-expression patterns. We demonstrated the correspondence of cell focus of risk genes to diseases’ location susceptibility and identified distinct cell-cell crosstalk and spatial heterogeneity of immune cell homing in different gut segments. Overall, a cross-spatiotemporal approach to transcriptomes at single-cell resolution revealed that the regional milieu of the human intestine determined cellular and molecular cues of immune surveillance, dictating gut homeostasis and disease.

Objective To compare the performances, tolerability and acceptability of mannitol and polyethylene glycol (PEG) as oral contrast agents in patients undergoing computed tomography enterography (CTE). Methods Patients aged 18‐75 years indicated for CTE were randomized to receive either mannitol or PEG as contrast agents. The coronal reconstructed images of each abdominal quadrant were assessed for maximum distention, proportion of distended bowel loops, presence of inhomogeneous contents and visibility of the small bowel wall. Overall subjective imaging quality assessment and patients’ tolerability and acceptability were recorded. Results Seventy patients were enrolled and randomized into two groups. In the per‐protocol analysis, no significant differences in imaging quality was found in bowel distention maximum diameter, wall visibility and intestinal homogeneity (all P > 0.05). The mean nausea score was lower in the mannitol group (0 [0‐0] vs 1.0 [0‐3.0], P

Background and aim Cresyl violet (CV) is a topical dye that allows simultaneous chromoendoscopy and in vivo confocal laser endomicroscopy in identification of neoplastic changes of the lower gastrointestinal tract without intravenous injection of fluorescein, but as yet no investigation has reported its application in the diagnosis of gastric intestinal metaplasia (GIM). This study aims to assess the feasibility as well as diagnosis accuracy of topical CV for in vivo diagnosis of GIM by using probe-based confocal laser endomicroscopy (pCLE). Methods In this prospective, open-label, feasibility study, 129 confocal videos from 22 patients with known GIM were analyzed and compared with corresponding histological images to establish the CV staining characteristics. In addition, 47 patients with known or suspected GIM were prospectively enrolled to evaluate the accuracy of this topical CV endomicroscopic imaging. Results Probe-based confocal laser endomicroscopy with topical CV enabled clear visualization of the goblet cells, absorptive cells, and intestinal villi of GIM. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of pCLE diagnosis of GIM on a per-location analysis was 93.01%, 91.95%, 93.51%, 86.96%, and 96.11%, respectively. The intraclass correlation coefficient for inter-observer agreement and mean kappa value for intra-observer agreement for the diagnosis of GIM was 0.82 and 0.87, respectively. Conclusions Topical CV enables real-time chromoendoscopy in conjunction with pCLE examination of the stomach and warrants accurate diagnosis of GIM. It may be an acceptable and potentially alternative dye for confocal imaging in the future.