Humans ; Endoscopic Ultrasound-Guided Fine Needle Aspiration ; Pancreatic Neoplasms ; Pancreas ; Needles

Objective:To construct a classification model for endoscopic ultrasonography (EUS) images of gastrointestinal stromal tumors (GISTs) and leiomyomas (LM) based on deep learning technology, and to verify its value for differential diagnosis.Methods:From October 2014 to October 2021, 69 patients of GISTs and 73 of LM who underwent EUS and were pathologically confirmed by surgery or endoscopic resection in the Second Affiliated Hospital of Soochow University were retrospectively studied. One clear EUS image with typical lesion was selected for each case. Using the hold-out method, the images of each disease were divided into the training set and the validation set according to the ratio of the number of images in the training set to the number of images in the validation set, which was 8∶2. Finally, 113 EUS images (55 GISTs and 58 LM) were used to form the training set, and 29 EUS images (14 GISTs and 15 LM) were used to form the validation set. The training set was used to train and optimize the deep learning model, and the validation set was used to verify the classification model. The main observation indicators included the sensitivity, the specificity, the positive predictive value, the negative predictive value and the accuracy of differential diagnosis.Results:The accuracy of the classification model established by Resnet 34 network structure in the differential diagnosis of GISTs and LM tended to be 0.89, better than the classification model established by Resnet 50 network structure (0.81). The sensitivity, the specificity, the positive predictive value, the negative predictive value and the accuracy of the classification model based on Resnet 34 network structure for differentiating EUS images in the validation set were 85.71% (12/14, 95% CI: 67.38%-100.00%), 93.33% (14/15, 95% CI: 80.71%-100.00%), 92.31% (12/13, 95% CI: 77.82%-100.00%), 87.50% (14/16, 95% CI: 71.30%-100.00%) and 89.66% (26/29, 95% CI: 78.57%-100.00%), respectively. Conclusion:It is feasible to use deep learning technology in the differential diagnosis of EUS images of GISTs and LM, which can provide auxiliary diagnostic opinions for clinicians. The deep learning model based on Resnet 34 network structure shows higher accuracy in the differential diagnosis of EUS images of GISTs and LM.

To compare the efficacy, safety and economic cost of endoscopic ultrasound (EUS)-guided puncture sclerotherapy and laparoscopic decapitation decompression for the renal cysts in the upper pole, data of patients with renal cysts in the upper pole who received EUS-guided puncture sclerotherapy (the EUS group, n=9) or laparoscopic decapitation decompression (the laparoscopy group, n=16) in the Second Affiliated Hospital of Soochow University from January 2021 to August 2022 were analyzed retrospectively. The effective rate, operation time, intraoperative blood loss, incidence of complications, hospital stay and treatment cost of the EUS group and the laparoscopy group were compared. Results showed that the effective rate was comparable in the EUS group and laparoscopy group (9/9 VS 14/16, P=0.520). The operation time was shorter (29.8±4.8 min VS 70.1±11.1 min, t=10.207, P<0.001), intraoperative blood loss less (0 mL VS 26.1±5.9 mL, t=13.089, P<0.001), postoperative hospital stay shorter (3.5±0.7 days VS 5.4±2.0 days, t=2.663, P=0.014), and total cost lower (10 547.85±2 388.19 yuan VS 15 316.09±5 352.45 yuan, t=2.517, P=0.019) in the EUS group compared with those in the laparoscopy group. There was no significant difference in the total hospital stay (8.1±2.0 days VS 9.3±3.1 days, t=1.019, P=0.319) or operation cost (3 946.79±490.82 yuan VS 3 860.18±857.42 yuan, t=-0.277, P=0.784) between the EUS group and laparoscopy group. There was 1 case of puncture bleeding, 1 case of hematuria, and 1 case of lumbago in the laparoscopy group, while no complication occurred in the EUS group. In conclusion, it is preliminarily believed that EUS-guided puncture sclerotherapy for renal cysts in the upper pole has similar clinical effects with higher safety, shorter postoperative hospital stay and lower total hospitalization cost compared with those of laparoscopic decapitation decompression, which is worth of clinical promotion.

Objective:To evaluate the preoperative localization value of endoscopic ultrasound guided fine needle tattooing (EUS-FNT) for laparoscopic distal pancreatectomy in pancreatic lesions with a maximum diameter ≤3 cm.Methods:From November 2017 to October 2022, at the Second Affiliated Hospital of Soochow University, the data of patients with pancreatic lesions ≤3 cm who underwent laparoscopic distal pancreatectomy were retrospectively analyzed. Eight patients who underwent EUS-FNT assisted laparoscopic distal pancreatectomy were included in the fine needle tattooing (FNT) combined laparoscopic group. And 14 patients who underwent simple laparoscopic distal pancreatectomy were taken as the simple laparoscopic group. The success rate and complications of EUS-FNT were observed. The differences in operation time, surgery-related complications and complete resection rate of lesions between the two groups were compared. Mann-Whitney U test and descriptive analysis were used for statistical analysis. Results:In the FNT combined laparoscopic group, the lesions of 4 cases were located in the pancreatic body and 4 cases in the pancreatic tail. In the simple laparoscopic group, the lesions of 4 cases were located in the pancreatic body and 10 cases in the pancreatic tail. There was a significant difference in lesion size between the two groups (14.5 mm (10.8 mm, 16.5 mm) vs. 27.0 mm (23.5 mm, 30.0 mm), Z=-3.09, P=0.001). In the FNT combined laparoscopic group, EUS-FNT was successfully performed in all 8 patients. The average time of laparoscopy after EUS-FNT was (98.4±8.8) min. The marks were clearly visible under the laparoscopic field of view, and no complications such as abdominal hemorrhage and hematoma were observed. Laparoscopic pancreaticocaudectomy was performed in 5 cases and pancreaticocaudectomy plus splenectomy in 3 cases. The median operation time was 192.5 min (176.3 min, 203.8 min). The amount of intraoperative bleeding was large in 2 patients and blood transfusion was needed. The lesions were one-time completely resected in all 8 patients. The postoperative pathology were 6 cases of pancreatic neuroendocrine neoplasm, 1 case of intraductal papillary mucinous neoplasm (IPMN), and 1 case of solid pseudopapilloma. In the simple laparoscopic group, laparoscopic pancreaticocaudectomy was performed in 2 cases and pancreaticocaudectomy plus splenectomy in 12 cases. The median operation time was 202.5 min (192.8 min, 235.0 min), which was longer than that of FNT combined laparoscopic group, but the difference was not statistically significant ( P>0.05). The amount of intraoperative bleeding was large in 2 patients and blood transfusion was needed. In 1 patient with pancreatic body lesions, no lesion was found in the specimen examination after the first pancreatectomy, and the lesions were completely resected after the second partial pancreatectomy. Active abdominal hemorrhage occurred in 1 patient on the second day after operation, and underwent interventional embolization for hemostasis. Two weeks after surgery, 1 patient was found to have a encapsulated fluid with a long diameter of 6 cm around the pancreas by computed tomography re-examination 2 weeks after surgery. The postoperative pathology were 5 cases of pancreatic neuroendocrine neoplasm, 2 cases of IPMN, 1 case of solid pseudopapilloma, 1 case of pancreatic cyst with glandular low-grade intraepithelial neoplasia, 1 case of ectopic spleen, and 4 cases of pancreatic ductal adenocarcinoma. Conclusion:EUS-FNT can effectively localize small pancreatic lesions before laparoscopic distal pancreatectomy, shorten the operation time and improve the complete resection rate under laparoscopy.

To evaluate the diagnostic value and safety of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) in patients with biopsy-negative malignant gastrointestinal stricture, a retrospective analysis was performed on data of patients whose computed tomography scan showed gastrointestinal malignant tumor with complete lumen stricture and endoscopic biopsy results showed negative, who underwent EUS-FNA in the Second Affiliated Hospital of Soochow University from July 2016 to January 2020. Perioperative complications, technical success rate and accuracy of EUS-FNA were analyzed. Eleven cases included in the study, including 7 males and 4 females, with mean age of 60.3 years. There were 8 esophageal strictures and 3 rectal strictures. All patients successfully underwent EUS-FNA, and malignant tumor was found in 10 cases and no tumor cell was found in 1 case. No complications were reported. EUS-FNA is a safe and valuable approach to diagnosing biopsy-negative malignant gastrointestinal strictures.

Endoscopic data of 108 upper gastrointestinal elevated lesions caused by vascular or hemangioma compression by endoscopic ultrasonography (EUS) at the Second Affiliated Hospital of Soochow University, Changshu No.1 People's Hospital, Kushan Hospital of Chinese Medicine and Traditional Chinese Medicine Hospital of Changshu from December 2010 to June 2019 were retrospectively summarized. The results showed that lesions were mainly located in the esophagus [50.9% (55/108)] and stomach [47.2% (51/108)], especially in the middle [40.0% (22/55)] and upper esophagus [36.4% (20/55)], body [66.7% (34/51)] and fundus of stomach [31.4% (16/51)], respectively. The major etiology included splenic artery and aneurysm compression [29.6% (32/108)], aortic compression [23.1% (25/108)], isolated esophageal venous aneurysm compression [13.9% (15/108)] and gastric submucosal vein and venous aneurysm compression [12.0% (13/108)], with diverse endoscopic presentation. The above results suggest that elevated lesions of upper gastrointestinal tract caused by blood vessels and hemangiomas are mostly due to external vascular pressure outside the lumen, but ectopic submucosal arteries and isolated phlebangioma are not uncommon. The lesions are widely distributed with different gastroscopic manifestations. EUS is important for definite diagnosis, and can be combined with color Doppler technique, CT plain scan and angiographic reconstruction if necessary.

Objective:A prospective, multicenter randomized controlled clinical research was conducted to explore the diagnostic value of the new optical staining technology for domestic endoscope, spectral focused imaging (SFI) and variable intelligent staining technology (VIST), for gastric precancerous lesions.Methods:Patients who intended to undergo gastroscopy between August 2020 and May 2021 were randomly divided into the white light group and the new optical staining group at the First Hospital of Hebei Medical University, Shanghai Tenth People's Hospital and the Second Affiliated Hospital of Soochow University. A sequential examination method was applied (white light to new optical staining or new optical staining to white light). The endoscopic diagnostic results and the detection results of Helicobacter pylori ( HP) of the two groups were recorded. At the same time, such five variables as gastric mucosal atrophy, intestinal metaplasia, fold enlargement, nodular gastritis and diffuse redness were evaluated for the risk of gastric cancer in the two groups. Results:A total of 419 cases were enrolled, including 208 cases in the white light group and 211 cases in the new optical staining group. Compared with pathological findings, the detection rates of gastric inflammation, atrophy, intestinal metaplasia, low-grade intraepithelial neoplasia, high-grade intraepithelial neoplasia and advanced cancer lesions in the white light group were 28.9%, 40.4%, 64.9%, 17.8%, 0.5% and 0.5% respectively; while those in the new optical staining group were 30.8%, 42.7%, 62.6%, 15.2%, 2.8% and 0.5%. There were no significant differences in the detection rates between the two groups ( P>0.05). Compared with pathology, the sensitivity, the specificity, the accuracy, the positive predictive value and the negative predictive value for gastric mucosal atrophy in the white light group were 92.9%, 61.3%, 74.0%, 61.9% and 92.7% respectively and those in the new optical staining group (SFI mode) were 94.4%, 64.5%, 77.3%, 66.4% and 94.0% respectively. The above 5 measures for gastric mucosal intestinal metaplasia were 68.1%, 72.6%, 69.7%, 82.1% and 55.2% in the white light group, and 87.1%, 89.9%, 88.2%, 93.5% and 80.7% in the new optical staining group (VIST mode), with significant difference between the two groups ( P<0.05). In terms of HP infection with 13C-urea breath test ( 13C-UBT) results as the gold standard, the above 5 measures were 90.2%, 84.3%, 87.4%, 86.8% and 88.2% in the white light group and 92.6%, 77.1%, 85.4%, 82.2% and 90.1% in the new optical staining group respectively. The proportion of high-risk gastric lesions in the new optical staining group was higher in cases of a gastric cancer risk score≥ 4 ( P<0.05). Conclusion:The new optical staining technology of domestic endoscopy has higher diagnostic value for gastric mucosal intestinal metaplasia. Gastroscopy is helpful for the detection of precancerous lesions with gastric cancer risk score as a tool. The new optical staining technology of domestic endoscopy is similar to imported endoscopy in diagnosing gastric precancerous lesions and HP infection, which is an effective means to detect gastric mucosal precancerous lesions.

Objective:To investigate the features of endoscopic ultrasonography in the diagnosis of malignant mediastinal and abdominal lymphadenopathy and to provide more evidence for endoscopic ultrasound-guided fine-needle aspiraiton (EUS-FNA).Methods:A case-control study was performed on 83 consecutive patients who underwent EUS in the Second Affiliated Hospital of Soochow University from September 2016 to February 2021. Lymph node properties were identified by pathological results of EUS-FNA and (or) surgery and follow-up for at least 6 months. According to the final diagnosis, patients were divided into malignant lymph node group ( n=56) and benign lymph node group ( n=27). Univariate analysis and multivariate logistic analysis were performed to identify independent risk factors for malignant lymphadenopathy in terms of EUS features. Results:Univariate analysis showed that the length of short axis, short-long axis ratio, shape, border, presence or absence of hilum, heterogeneous echo, and the growth pattern of lymph node were risk factors for malignant lymph nodes ( P<0.10). Multivariate logistic regression analysis showed that short axis>10 mm ( P=0.021, OR=9.751, 95% CI: 1.407-57.573), clear border ( P=0.009, OR=20.587, 95% CI: 2.149-197.251), absence of hilum ( P=0.019, OR=28.502, 95% CI: 1.725-470.864), nodal matting ( P=0.004, OR=45.539, 95% CI: 3.429-604.822), partial nodal fusion ( P=0.004, OR=50.012, 95% CI: 3.497-715.266) were independent risk factors for malignant mediastinal and abdominal lymph nodes. Conclusion:EUS is useful to differentiate the lymph node properties in the mediastinal or abdominal cavity. Short axis>10 mm, clear border, absence of hilum, nodal matting and partial nodal fusion are high-risk EUS features of malignant mediastinal or abdominal lymphadenopathy, where priority should be given to EUS-FNA.

To evaluate the diagnostic value of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) for portal vein tumor thrombus, data of patients with digestive system malignant tumors combined with portal vein tumor thrombus diagnosed by CT or magnetic resonance imaging who underwent EUS-FNA from April 2015 to July 2020 in the Second Affiliated Hospital of Soochow University were collected. A total of 7 patients were included, with 2 cases of primary hepatocellular carcinoma, 3 cases of primary pancreatic carcinoma and 2 cases of primary gastric cancer. EUS-FNA was successfully performed in 7 patients with portal vein embolus. Pathological examination of portal vein embolus showed 5 cases of malignant tumor. No tumor cell was found in 2 cases. There were no complications such as local hematoma, abdominal hemorrhage or infection in all patients. EUS-FNA is safe and effective for patients diagnosed as having malignant tumors with portal vein embolus.

Liver injury caused by viral and non-viral factors is an important stage of chronic liver disease, and the pathogenesis of liver injury is still a research hotspot. With the deepening of the research on gut microbiota, substantial evidence indicates that gut microbiota participates in the development and progression of liver injury, and it has been confirmed that Akkermansia muciniphila (Akk) has a beneficial effect against liver injury. This article summarizes the role and potential mechanism of Akk in immune-mediated liver injury, alcoholic liver disease, and nonalcoholic fatty liver disease. It is believed that Akk may provide new directions and choices for the prevention and treatment of liver injury.