Abstract:Background and Objectives: In the recent years, EUS is one of the routine procedures in the diagnosis of pancreatic diseases. EUS-guided needle-based confocal laser endomicroscopy (nCLE) is a novel minimally invasive imaging technique in diagnosis of pancreatic diseases. The pilot researches provided us some preliminary findings and conclusions with small samples, low rate of pathological correspondence. The aim of this current study was to evaluate the diagnostic efficacy of EUS-guided nCLE in solid pancreatic lesions (SPLs) and pancreatic cystic lesions (PCLs) based on large samples. The date was obtained on nCLE imaging findings and high rate of correlation with pathology. Material and Methods: Patients enrolled in the study were underwent EUS-nCLE to achieve the nCLE images and diagnosis. Comparing with the final diagnosis, including surgical histopathological results or cyto-/histopathology through FNA, the efficacy and accuracy of nCLE in diagnosis in solid and cystic pancreatic lesions were evaluated. In other cases, clinical diagnoses were achieved based on the combination with clinical history, image findings and fluid analysis and cytology, by 3 independent committee members strongly agreed with a concordant diagnosis. Results: Totally 172 patients were enrolled into the study. The overall rate of final diagnosis was about 65% while 50% in cystic lesion. The mean sensitivity, specificity, negative predictive value, positive predictive value and accuracy of the nCLE in diagnosis of PDAC is 90.3%, 89.5%, 93.3%, 85.0% and 90.0% respectively. The efficacy and accuracy of pancreatic cystic lesions were very satisfying and some additional nCLE signs were found, including "black aggregates of cells, forming as gland-like structure, surrounding by fibro and vessels" in neuroendocrine tumors (NETs); "black columnar protrusions near vascular area" in the pseudopapillary solid tumor (SPT); macrophage in tuberculosis (TB) and small aggregate of black regular cells maybe corresponds to ovarian-like stroma in mucinous cystadenoma (MCN). In the study, 20 (11.6%) patients suffered complications, including symptomatic (5.2%) and asymptomatic (6.4%). Conclusions: nCLE observation could improve characterization of indeterminate cysts, or confirm the EUS impression, when cytological confirmation is missing. The technique may deliver information to better guide our clinical decisions.

A convolutional neural network-based system for identifying neuroendocrine neoplasms and multiple types of lesions in the pancreas using EUS (with videos)

Utilisation of artificial intelligence for the development of an EUS-convolutional neural network model trained to enhance the diagnosis of autoimmune pancreatitis

Deep learning and automatic differentiation of pancreatic lesions in endoscopic ultrasound – a transatlantic study

Artificial Intelligence-Based Diagnosis of Standard Endoscopic Ultrasonography Scanning Sites in the Biliopancreatic System: A Multicenter Retrospective Study

Application of endoscopic ultrasonography for detecting esophageal lesions based on convolutional neural network.

Contrast-enhanced harmonic endoscopic ultrasound (CH-EUS) MASTER: A novel deep learning-based system in pancreatic mass diagnosis

A deep learning model using hyperspectral image for EUS‐FNA cytology diagnosis in pancreatic ductal adenocarcinoma

Endoscopic ultrasound diagnosis system based on deep learning in images capture and segmentation training of solid pancreatic masses

Convolutional neural network-based system for endocytoscopic diagnosis of early gastric cancer

Enhancing detection of various pancreatic lesions on endoscopic ultrasound through artificial intelligence: a basis for computer‐aided detection systems

The Application of Endoscopic Ultrasound-Guided Needle-Based Confocal Laser Endomicroscopy in Diagnosis of Pancreatic Neoplasms

Appraisal of EUS-guided needle-based confocal laser endomicroscopy in the diagnosis of pancreatic lesions: A single Chinese center experience

Convolutional neural network-based artificial intelligence for the diagnosis of early esophageal cancer based on endoscopic images: A meta-analysis

High Accuracy of Convolutional Neural Network for Evaluation of Helicobacter Pylori Infection Based on Endoscopic Images: Preliminary Experience

Artificial Intelligence Efficiently Predicts Gastric Lesions, Helicobacter Pylori Infection and Lymph Node Metastasis Upon Endoscopic Images.

A Deep Learning-Based Segmentation System for Rapid Onsite Cytologic Pathology Evaluation of Pancreatic Masses: A Retrospective, Multicenter, Diagnostic Study

AI-assisted Identification of Intrapapillary Capillary Loops in Magnification Endoscopy for Diagnosing Early-Stage Esophageal Squamous Cell Carcinoma: a Preliminary Study

Deep learning-based pancreas segmentation and station recognition system in EUS: development and validation of a useful training tool (with video)

The Diagnostic Value of Endoscopic Ultrasonography for Pancreatic Neuroendocrine Tumors

Convolutional neural network for the diagnosis of early gastric cancer based on magnifying narrow band imaging

Deep learning for real-time auxiliary diagnosis of pancreatic cancer in endoscopic ultrasonography