Search

Simple Summary: Surgical therapy is critical to pancreatic cancer survival. The segmentation of pancreatic cancer in endoscopic ultrasonography (EUS) images can provide critical characteristics of the pancreatic cancer for surgical therapy. However, EUS has high operator dependency, and it requires a considerable level of experience and competency to stage pancreatic cancer. Deep learning approaches have been used on EUS images, but there have been no studies on the segmentation of pancreatic cancer. The purpose of this study is to segment pancreatic cancer from EUS images using a neural network model with deep attention features, called DAF-Net. The significance of this study lies in the successful segmentation performance of the pancreatic cancer using the DAF-Net, regardless of the size and location of the cancer, and its usefulness in preoperative planning. Endoscopic ultrasonography (EUS) plays an important role in diagnosing pancreatic cancer. Surgical therapy is critical to pancreatic cancer survival and can be planned properly, with the characteristics of the target cancer determined. The physical characteristics of the pancreatic cancer, such as size, location, and shape, can be determined by semantic segmentation of EUS images. This study proposes a deep learning approach for the segmentation of pancreatic cancer in EUS images. EUS images were acquired from 150 patients diagnosed with pancreatic cancer. A network with deep attention features (DAF-Net) is proposed for pancreatic cancer segmentation using EUS images. The performance of the deep learning models (U-Net, Attention U-Net, and DAF-Net) was evaluated by 5-fold cross-validation. For the evaluation metrics, the Dice similarity coefficient (DSC), intersection over union (IoU), receiver operating characteristic (ROC) curve, and area under the curve (AUC) were chosen. Statistical analysis was performed for different stages and locations of the cancer. DAF-Net demonstrated superior segmentation performance for the DSC, IoU, AUC, sensitivity, specificity, and precision with scores of 82.8%, 72.3%, 92.7%, 89.0%, 98.1%, and 85.1%, respectively. The proposed deep learning approach can provide accurate segmentation of pancreatic cancer in EUS images and can effectively assist in the planning of surgical therapies. [ABSTRACT FROM AUTHOR]

Background and aims: Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) is frequently used for the preoperative histologic diagnosis of pancreatic cancer. However, debate continues regarding the clinical merits of preoperative EUS-FNA for the management of resectable pancreatic cancer. We aimed to evaluate the benefits and safety of preoperative EUS-FNA for resectable distal pancreatic cancer. Methods: The medical records of 304 consecutive patients with suspected distal pancreatic cancer who underwent EUS-FNA were retrospectively reviewed to evaluate the clinical benefits of preoperative EUS-FNA. We also reviewed the medical records of 528 patients diagnosed with distal pancreatic cancer who underwent distal pancreatectomy with or without EUS-FNA. The recurrence rates and cancer-free survival periods of patients who did or did not undergo preoperative EUS-FNA were compared. Results: The diagnostic accuracy of preoperative EUS-FNA was high (sensitivity, 87.5%; specificity, 100%; positive predictive value 100%; accuracy, 90.7%; negative predictive value, 73.8%). Among patients, 26.7% (79/304) avoided surgery based on the preoperative EUS-FNA findings. Of the 528 patients who underwent distal pancreatectomy, 193 patients received EUS-FNA and 335 did not. During follow-up (median 21.7 months), the recurrence rate was similar in the two groups (EUS-FNA, 72.7%; non-EUS-FNA, 75%; P = 0.58). The median cancer-free survival was also similar (P = 0.58); however, gastric wall recurrence was only encountered in the patients with EUS-FNA (n = 2). Conclusion: Preoperative EUS-FNA is not associated with increased risks of cancer-specific or overall survival. However, clinicians must consider the potential risks of needle tract seeding, and care should be taken when selecting patients. [ABSTRACT FROM AUTHOR]

Simple Summary: Pancreatic cancer (PC) is one of the most lethal cancers; caused by family history, obesity, diabetes, and smoking, it has a 2–9% five-year survival rate. However, patients diagnosed by endoscopic ultrasound (EUS) already have an advanced stage of PC, indicating the difficulty of surgical resection. Recently, laser ablative treatment with a diffusing applicator has been proven to be feasible for treating advanced PC. Despite the potential application for treating PC, further evaluation of acute and chronic tissue responses are essential to determine the efficacy and safety of diffusing applicator under EUS guidance. In this study, EUS-guided diffusing applicator-assisted laser ablation was evaluated to quantify the extent of the tissue response after the delivery of various energy levels. Endoscopic ultrasound (EUS)-guided cylindrical interstitial laser ablation (CILA) procedures can be used to treat unresectable pancreatic cancer (PC). The aim of this study was to investigate the acute responses of pancreatic tissue after EUS-guided CILA in vivo in porcine models. Eight pigs were tested to compare the effects of different energy levels on pancreatic tissue ablation. A 1064 nm laser system was used to deliver 5 W through a diffusing applicator. The EUS-guided CILA was performed under four different energies: 200, 400, 600, and 800 J. Three days after the experiments, histological analysis was performed. The CILA consistently generated circular coagulated necrosis (CN) in the cross-sectioned pancreatic tissue. The ablation diameter was linearly dependent on the total energy delivery. The area of the CN initially increased with total energy delivery but became saturated at 600 J. The width of the degenerative parenchyma (DP) in the native tissue beyond the CN region increased with the total energy up to 600 J, and then decreased afterward. EUS-guided CILA can be a feasible approach for treating PC. Further animal studies will investigate the chronic responses of the pancreatic tissue to examine the efficacy and safety of the proposed method for clinical translation. [ABSTRACT FROM AUTHOR]

Background: Endoscopic papillary large balloon dilation (EPLBD) without prior endoscopic sphincterotomy (EST) produces excellent outcomes for the treatment of large common bile duct (CBD) stones. However, it remains unclear how the outcomes of EPLBD alone compare with those of EPLBD with EST. In this study, we assessed the safety and therapeutic outcomes of EPLBD with vs. without EST for the removal of large bile duct stones.Methods: This prospective, multicenter study was conducted on 200 patients with bile duct stones of ≥ 10 mm in diameter. Patients were randomly assigned to an EPLBD alone group (n = 100) or an EPLBD with EST group (n = 100). These two groups were compared with respect to overall procedure-related adverse events, overall stone removal success rate, number of endoscopic sessions required for complete stone removal, need for mechanical lithotripsy, and total procedure time.Results: The incidence of adverse events was not significantly different between the groups (EPLBD alone vs. EPLBD with EST: overall adverse events 6 % vs. 4 %, P = 0.75; pancreatitis 1 % vs. 3 %, P = 0.62). Overall success (P = 0.35), initial success (P = 0.28), and the need for mechanical lithotripsy (P = 0.39) were also similar between groups. Median total procedure time tended to be greater in the EPLBD alone group (20.5 minutes) than in the EPLBD with EST group (18 minutes; P = 0.08).Conclusion: The therapeutic outcomes and adverse events of EPLBD alone for the removal of large bile duct stones were comparable to those of EPLBD with EST. [ABSTRACT FROM AUTHOR]

Background and Aim Radiofrequency ablation (RFA) and ethanol ablation are accepted methods of tissue destruction for treating cystic tumors. The aim of the present study was to evaluate the feasibility, efficacy, and safety of endoscopic ultrasound (EUS)-guided ablation using normal porcine gallbladders as a substitute model for branch-duct intraductal papillary mucinous neoplasms (BD-IPMN). Methods Six adult mini pigs were included in this prospective study. EUS-guided RFA with or without ethanol ablation of the gallbladder was carried out using a prototype 18-gauge endoscopic RFA electrode, and RFA and ethanol ablation were done in a single session. Outcomes were assessed in terms of macroscopic and microscopic evaluations of the treated gallbladders. Results The prototype RFA electrode was used for ablation of the mucosa, aspiration of the internal contents of the gallbladder, and ethanol injection. RFA plus ethanol lavage resulted in a greater mean percentage of denuded gallbladder mucosa (97.5%) than RFA alone (55.8%). Effects of ablation extending to the cystic duct and partially denuded cystic duct mucosae were detected in five of the six pigs. There were no major procedure-related adverse events. Conclusions EUS-guided ablation of the gallbladder mucosa was feasible and effective in the porcine model. Additional studies will be required to fully assess the risk of procedure-related damage to the main pancreatic duct before using this technique in a clinical setting. [ABSTRACT FROM AUTHOR]

Mucinous cystic neoplasms (MCN) and serous cystic neoplasms (SCN) account for a large portion of solitary pancreatic cystic neoplasms (PCN). In this study we implemented a convolutional neural network (CNN) model using ResNet50 to differentiate between MCN and SCN. The training data were collected retrospectively from 59 MCN and 49 SCN patients from two different hospitals. Data augmentation was used to enhance the size and quality of training datasets. Fine-tuning training approaches were utilized by adopting the pre-trained model from transfer learning while training selected layers. Testing of the network was conducted by varying the endoscopic ultrasonography (EUS) image sizes and positions to evaluate the network performance for differentiation. The proposed network model achieved up to 82.75% accuracy and a 0.88 (95% CI: 0.817–0.930) area under curve (AUC) score. The performance of the implemented deep learning networks in decision-making using only EUS images is comparable to that of traditional manual decision-making using EUS images along with supporting clinical information. Gradient-weighted class activation mapping (Grad-CAM) confirmed that the network model learned the features from the cyst region accurately. This study proves the feasibility of diagnosing MCN and SCN using a deep learning network model. Further improvement using more datasets is needed. [ABSTRACT FROM AUTHOR]