Purpose To automatically and efficiently segment the lesion area of the colonoscopy polyp image, a polyp segmentation method has been presented. Methods An ensemble model of pretrained convolutional neural networks was proposed, using Unet‐VGG, SegNet‐VGG, and PSPNet. Firstly, the Unet‐VGG is obtained by the first 10 layers of VGG16 as the contraction path of the left half of the Unet. Then, the SegNet‐VGG is acquired by fine‐tuned transfer learning VGG16, using the first 13 layers of VGG16 as the encoder of the SegNet and combined the original decoder of the SegNet. By adjusting the input size of the Unet‐VGG, SegNet‐VGG, and PSPNet, the preprocessed data can be correctly fed to the three network models. The three models are used as the basic trainer to train and segment the datasets. Based on the ensemble learning algorithm, the weight voting method is used to ensemble the segmentation results corresponding to single basic trainer. Results Both IoU and DICE similarity score were used to evaluate the segmentation quality for cvc300 with 300 images, CVC‐ClinicDB with 612 images, and ETIS‐LaribPolypDB with 196 images. From the experimental results, the IoU and DICE obtained by the proposed method for the cvc300 datasets can reach up to 96.16% and 98.04%, respectively, the IoU and DICE for the CVC‐ClinicDB datasets can reach up to 96.66% and 98.30%, respectively, whereas the IoU and DICE for the ETIS‐LaribPolypDB datasets can reach up to 96.95% and 98.45%, respectively. Evaluation of the IoU and DICE in our methods shows higher accuracy than previous methods. Conclusions The experimental results show that the proposed method improved correspondingly in IoU and DICE compared to a single basic trainer. The range of improvement is 1.98%–6.38%. The proposed ensemble learning succeeds in automatic polyp segmentation, which potentially helps to establish more polyp datasets.
Although endoscopic ultrasound (EUS)-guided transmural drainage of pancreatic fluid collections with metal stents is generally preferred over plastic stents, its superiority among different types of metal stents has not yet been well studied. We conducted this study to compare clinical outcomes and complications of a novel self-expanding biflanged metal stent (BFMS) and a traditional-shaped tubular metal stent (TMS) in treating pancreatic pseudocyst (PPC).This was a retrospective analysis on consecutive patients with PPC underwent EUS-guided transmural drainage with either TMS or BFMS in a single tertiary center with expertise in management of complex biliary and pancreatic problems. The technical and functional success rate, reintervention, complications, and recurrence rate were evaluated.From September 2013 to January 2018, 125 patients (66.4% male, median age 47 years) underwent EUS-guided transmural drainage for PPC. Among them, 49 used TMS and 76 used BFMS. All patients met the inclusion criteria that cyst diameter was >6 cm or the distance between cyst and stomach wall was shorter than 1 cm. There was no difference in technical success (98% vs 97.4%, P = 1.0) or functional success rate (87.8% vs 92.1%, P = .54) using 2 types of metal stents. However, more procedure related complications occurred in TMS than in BFMS group. TMS group had a much higher migration rate than BFMS group (14.6% vs 0, P = .001), even though there was no significant difference in bleeding, infection, or death rate between 2 groups. With similar clinical outcomes, TMS group required more additional plastic stent placement than BFMS group for better drainage.TMS and BFMS placement can both be considered as methods of endoscopic transmural PPC drainage with equal efficacy, whereas BFMS could be preferred for fewer complications or less need of additional plastic stent placement.
Acute pancreatitis (AP) usually causes acute lung injury, which is also known as acute pancreatitis associated lung injury (APALI). This study aimed to investigate whether captopril pretreatment was able to protect lung against APALI via inhibiting angiotensin II (Ang II) production and suppressing Rho/ROCK (Rho kinase) pathway in rats. Severe AP (SAP) was introduced to rats by bile-pancreatic duct retrograde injection of 5% sodium taurocholate. Rats were randomly divided into three groups. In the sham group, sham operation was performed; in the SAP group, SAP was introduced; in the pre-cpl + SAP group, rats were intragastrically injected with 5 mg/kg captopril 1 hour prior to SAP induction. Pathological examination of the lung and pancreas, evaluation of pulmonary vascular permeability by wet/dry ratio and Evans Blue staining, detection of serum amylase, Western blot assay for Ang II receptor type 1 (AT1), RhoA, ROCK (Rho kinase), and MLCK (myosin light chain kinase) were performed after the animals were sacrificed at 24 hours. After the surgery, characteristic findings of pancreatitis were observed, accompanied by lung injury. The serum amylase, Ang II, and lung expression of AT1, RhoA, ROCK, and MLCK increased dramatically in SAP rats. However, captopril pretreatment improved the histological changes, reduced the pathological score of the pancreas and lung, inhibited serum amylase and Ang II production, and decreased expression of AT1, RhoA, ROCK, and MLCK in the lung. These findings suggest that captopril pretreatment is able to protect the lung against APALI, which is, at least partially, related to the inhibition of Ang II production and the suppression of the Rho/ROCK pathway.
Background and Aim:Brush cytology is widely applied to diagnosis indeterminate biliary stricture but suffer from low sensitivity. Changes in DNA content are a character of malignant cell and can be detected by DNA image cytometry (DNA-ICM). The study aimed to estimate the value of routine cytology (RC), DNA-ICM, and their combination in diagnosing indeterminate biliary strictures. Methods: A total of 362 patients who underwent both RC and DNA-ICM tests were analysed. Their results were retrospectively applied to final diagnoses. Diagnostic values were compared among RC, DNA-ICM, and their combination based on the location of strictures. Results: The DNA-ICM and combination of two methods had higher diagnostic accuracy than RC in all strictures (63.3% vs 42.3%, P < 0.001, 64.36% vs 42.3%, P < 0.001) and in distal strictures (65.36% vs 42.81%, P < 0.001, 66.01% vs 42.81%, P < 0.001). But in proximal strictures, DNA-ICM showed no superior (51.8% vs 42.81%, P = 0.184). Combination of two methods was not fully significant superior to RC in proximal strictures (55.36% vs 39.29%, P = 0.089). After classification of "suspicious for malignancy" as positive for malignancy, the diagnostic accuracy of DNA-ICM was still higher than that of RC in all strictures (63.3% vs 51.9%, P = 0.002) and in distal strictures (65.36% vs 52.29%, P = 0.001). Combination of two methods was no superior to DNA-ICM alone (64.36% vs 63.3%, P = 0.757). The utilization of DNA-ICM was more accurate in distal strictures than in proximal strictures (65.36% vs 51.8%, P = 0.017). Conclusion: DNA-ICM is an objective and effective addition tool with RC, especially in distal strictures. The combination of DNA-ICM and RC showed no superior to DNA-ICM alone but could improve diagnostic accuracy to RC in proximal strictures although not fully significant. ) 0.738 (0.625-0.853) † Suspicious cytology findings were considered indications of malignancy.AUC, area under curve; DNA-ICM, DNA image cytometry; NPV, negative predictive value; PPV, positive predictive value; SE, sensitivity; SP, specificity.
BACKGROUND Early detection of advanced cystic mucinous neoplasms [(A-cMNs), defined as high-grade dysplasia or malignancy] of the pancreas is of great significance. As a simple and feasible detection method, serum tumor markers (STMs) may be used to predict advanced intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs). However, there are few studies on the usefulness of STMs other than carbohydrate antigen (CA) 19-9 for early detection of A-cMNs. AIM To study the ability of five STMs-CA19-9, carcinoembryonic antigen (CEA), CA125, CA724, and CA242 to predict A-cMNs and distinguish IPMNs and MCNs. METHODS We mainly measured the levels of each STM in patients pathologically diagnosed with cMNs. The mean levels of STMs and the number of A-cMN subjects with a higher STM level than the cutoff were compared respectively to identify the ability of STMs to predict A-cMNs and distinguish MCNs from IPMNs. A receiver operating characteristic curve with the area under curve (AUC) was also created to identify the performance of the five STMs. RESULTS A total of 187 patients with cMNs were identified and 72 of them showed A-cMNs. We found that CA19-9 exhibited the highest sensitivity (SE) (54.2%) and accuracy (76.5%) and a moderate ability (AUC = 0.766) to predict A-cMNs. In predicting high-grade dysplasia IPMNs, the SE of CA19-9 decreased to 38.5%. The ability of CEA, CA125, and CA724 to predict A-cMNs was low (AUC = 0.651, 0.583, and 0.618, respectively). The predictive ability of CA242 was not identified. The combination of STMs improved the SE to 62.5%. CA125 may be specific to the diagnosis of advanced MCNs. CONCLUSION CA19-9 has a moderate ability, and CEA, CA125, and CA724 have a low ability to predict A-cMNs. The combination of STM testing could improve SE in predicting A-cMNs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.