Endoscopic Ultrasound Imaging Research Articles

Construction and validation of an endoscopic ultrasonography-based ultrasomics nomogram for differentiating pancreatic neuroendocrine tumors from pancreatic cancer

ObjectivesTo develop and validate various ultrasomics models based on endoscopic ultrasonography (EUS) for retrospective differentiating pancreatic neuroendocrine tumors (PNET) from pancreatic cancer.MethodsA total of 231 patients, comprising 127 with pancreatic cancer and 104 with PNET, were retrospectively enrolled. These patients were randomly divided into either a training or test cohort at a ratio of 7:3. Ultrasomics features were extracted from conventional EUS images, focusing on delineating the region of interest (ROI) for pancreatic lesions. Subsequently, dimensionality reduction of the ultrasomics features was performed by applying the Mann-Whitney test and least absolute shrinkage and selection operator (LASSO) algorithm. Eight machine learning algorithms, namely logistic regression (LR), light gradient boosting machine (LightGBM), multilayer perceptron (MLP), random forest (RF), extra trees, k nearest neighbors (KNN), support vector machine (SVM), and extreme gradient boosting (XGBoost), were employed to train prediction models using nonzero coefficient features. The optimal ultrasomics model was determined using a ROC curve and utilized for subsequent analysis. Clinical-ultrasonic features were assessed using both univariate and multivariate logistic regression. An ultrasomics nomogram model, integrating both ultrasomics and clinical-ultrasonic features, was developed.ResultsA total of 107 EUS-based ultrasomics features were extracted, and 6 features with nonzero coefficients were ultimately retained. Among the eight ultrasomics models based on machine learning algorithms, the RF model exhibited superior performance with an AUC= 0.999 (95% CI 0.9977 - 1.0000) in the training cohort and an AUC= 0.649 (95% CI 0.5215 - 0.7760) in the test cohort. A clinical-ultrasonic model was established and evaluated, yielding an AUC of 0.999 (95% CI 0.9961 - 1.0000) in the training cohort and 0.847 (95% CI 0.7543 - 0.9391) in the test cohort. Subsequently, the ultrasomics nomogram demonstrated a significant improvement in prediction accuracy in the test cohort, as evidenced by an AUC of 0.884 (95% CI 0.8047 - 0.9635) and confirmed by the Delong test. The calibration curve and decision curve analysis (DCA) depicted this ultrasomics nomogram demonstrated superior accuracy. They also yielded the highest net benefit for clinical decision-making compared to alternative models.ConclusionsA novel ultrasomics nomogram was proposed and validated, that integrated clinical-ultrasonic and ultrasomics features obtained through EUS, aiming to accurately and efficiently identify pancreatic cancer and PNET.

Radiomics analysis for prediction and classification of submucosal tumors based on gastrointestinal endoscopic ultrasonography.

To identify and classify submucosal tumors by building and validating a radiomics model with gastrointestinal endoscopic ultrasonography (EUS) images. A total of 144 patients diagnosed with submucosal tumors through gastrointestinal EUS were collected between January 2019 and October 2020. There are 1952 radiomic features extracted from each patient's EUS images. The statistical test and the customized least absolute shrinkage and selection operator regression were used for feature selection. Subsequently, an extremely randomized trees algorithm was utilized to construct a robust radiomics classification model specifically tailored for gastrointestinal EUS images. The performance of the model was measured by evaluating the area under the receiver operating characteristic curve. The radiomics model comprised 30 selected features that showed good discrimination performance in the validation cohorts. During validation, the area under the receiver operating characteristic curve was calculated as 0.9203 and the mean value after 10-fold cross-validation was 0.9260, indicating excellent stability and calibration. These results confirm the clinical utility of the model. Utilizing the dataset provided curated from gastrointestinal EUS examinations at our collaborating hospital, we have developed a well-performing radiomics model. It can be used for personalized and non-invasive prediction of the type of submucosal tumors, providing physicians with aid for early treatment and management of tumor progression.

A novel artificial intelligence-based endoscopic ultrasonography diagnostic system for diagnosing the invasion depth of early gastric cancer.

We developed an artificial intelligence (AI)-based endoscopic ultrasonography (EUS) system for diagnosing the invasion depth of early gastric cancer (EGC), and we evaluated the performance of this system. A total of 8280 EUS images from 559 EGC cases were collected from 11 institutions. Within this dataset, 3451 images (285 cases) from one institution were used as a development dataset. The AI model consisted of segmentation and classification steps, followed by the CycleGAN method to bridge differences in EUS images captured by different equipment. AI model performance was evaluated using an internal validation dataset collected from the same institution as the development dataset (1726 images, 135 cases). External validation was conducted using images collected from the other 10 institutions (3103 images, 139 cases). The area under the curve (AUC) of the AI model in the internal validation dataset was 0.870 (95% CI: 0.796-0.944). Regarding diagnostic performance, the accuracy/sensitivity/specificity values of the AI model, experts (n = 6), and nonexperts (n = 8) were 82.2/63.4/90.4%, 81.9/66.3/88.7%, and 68.3/60.9/71.5%, respectively. The AUC of the AI model in the external validation dataset was 0.815 (95% CI: 0.743-0.886). The accuracy/sensitivity/specificity values of the AI model (74.1/73.1/75.0%) and the real-time diagnoses of experts (75.5/79.1/72.2%) in the external validation dataset were comparable. Our AI model demonstrated a diagnostic performance equivalent to that of experts.

Revolutionizing the Pancreatic Tumor Diagnosis: Emerging Trends in Imaging Technologies: A Systematic Review

Background and Objectives: The pancreas, ensconced within the abdominal cavity, requires a plethora of sophisticated imaging modalities for its comprehensive evaluation, with ultrasonography serving as a primary investigative technique. A myriad of pancreatic pathologies, encompassing pancreatic neoplasia and a spectrum of inflammatory diseases, are detectable through these imaging strategies. Nevertheless, the intricate anatomical confluence and the pancreas’s deep-seated topography render the visualization and accurate diagnosis of its pathologies a formidable endeavor. The objective of our paper is to review the best diagnostic imagistic tools for the pancreas. Materials and Methods: we have gathered several articles using Prisma guidelines to determine the best imagistic methods. The imperative of pancreatic scanning transcends its diagnostic utility, proving to be a pivotal element in a multitude of clinical specialties, notably surgical oncology. Within this domain, multidetector computed tomography (MDCT) of the pancreas holds the distinction of being the paramount imaging modality, endorsed for its unrivaled capacity to delineate the staging and progression of pancreatic carcinoma. In synergy with MDCT, there has been a notable advent of avant-garde imaging techniques in recent years. These advanced methodologies, including ultrasonography, endoscopic ultrasonography, contrast-enhanced ultrasonography, and magnetic resonance imaging (MRI) conjoined with magnetic resonance cholangiopancreatography (MRCP), have broadened the horizon of tumor characterization, offering unparalleled depth and precision in oncological assessment. Other emerging diagnostic techniques, such as elastography, also hold a lot of potential and promise for the future of pancreatic imaging. Fine needle aspiration (FNA) is a quick, minimally invasive procedure to evaluate lumps using a thin needle to extract tissue for analysis. It is less invasive than surgical biopsies and usually performed as an outpatient with quick recovery. Its accuracy depends on sample quality, and the risks include minimal bleeding or discomfort. Results, guiding further treatment, are typically available within a week. Elastography is a non-invasive medical imaging technique that maps the elastic properties and stiffness of soft tissue. This method, often used in conjunction with ultrasound or MRI, helps differentiate between hard and soft areas in tissue, providing valuable diagnostic information. It is particularly useful for assessing liver fibrosis, thyroid nodules, breast lumps, and musculoskeletal conditions. The technique is painless and involves applying gentle pressure to the area being examined. The resulting images show tissue stiffness, indicating potential abnormalities. Elastography is advantageous for its ability to detect diseases in early stages and monitor treatment effectiveness. The procedure is quick, safe, and requires no special preparation, with results typically available immediately. Results: The assembled and gathered data shows the efficacy of various techniques in discerning the nature and extent of neoplastic lesions within the pancreas. Conclusions: The most common imaging modalities currently used in diagnosing pancreatic neoplasms are multidetector computed tomography (MDCT), endoscopic ultrasound (EUS), and magnetic resonance imaging (MRI), alongside new technologies, such as elastography.

Efficacy and histologic frequencies of endoscopic ultrasonography‐guided tissue acquisition using conventional fine‐needle aspiration needles for gastric subepithelial hypoechoic mass

AbstractIntroduction/PurposeFor gastric subepithelial lesions (GSELs) showing a hypoechoic mass (HM) on endoscopic ultrasonography (EUS) imaging, the utility of EUS‐guided tissue acquisition using conventional fine‐needle aspiration needles (EUS‐TA‐CFNAN) and the frequency of histological types remain unclear. This study aimed to examine this issue.MethodsThis prospective observational study enrolled 291 consecutive patients who underwent EUS‐TA‐CFNAN for GSELs showing an HM (GSELHM) on EUS imaging. Immunohistochemical analysis was performed for all EUS‐TA‐CFNAN and surgically resected specimens. The main outcome measures were the technical results of EUS‐TA‐CFNAN and the frequency of histological types in GSELHM.ResultsThe endoscopic ultrasound‐guided tissue acquisition using conventional fine‐needle aspiration needle diagnosis rate for GSELHM was 80.1% (95% confidence interval [CI]: 75.0–84.5, 233/291). It was significantly lower for antrum (P = 0.004) and lesions smaller than 2 cm (P = 0.003). There were no adverse events. The immunohistochemical diagnoses of EUS‐TA‐CFNAN included 149 cases of gastrointestinal stromal tumour (GIST) (51.2%), 48 cases of leiomyoma (16.5%), 11 cases of schwannoma (3.8%), 8 cases of the ectopic pancreas (2.7%), 5 cases of subepithelial lesion like cancer (1.7%), 12 cases of other lesions (4.1%), and 58 cases of undiagnosable lesions (19.9%). The frequency of malignant or potentially malignant tumour in GSELHM was 55.0% (95% CI: 49.1–60.8, 160/291). Surgery was performed in 149 patients according to the conclusive EUS‐TA‐CFNAN results, in which the diagnostic accuracy of EUS‐TA‐CFNAN was 97.3% (95% CI: 94.7–99.9, 145/149).ConclusionThe use of EUS‐TA‐CFNAN for GSELHMs is safe and accurate. Gastric subepithelial lesions showing a hypoechoic mass have a reasonably high possibility of containing malignant or potentially malignant tumours, including GISTs.

Extraperitoneal lateral pelvic sidewall excision: a novel rectal-sparing approach for lateral locally recurrent rectal cancer.

Locally recurrent rectal cancer (LRRC) involving the lateral pelvic sidewall requires a complex approach to maximize the likelihood of R0 resection, which is the only predictor of survival. The purpose of this report is to describe a novel technique to resect a localized lateral pelvic sidewall LRRC. A 63-year-old male patient was referred for a 15-mm LRRC near the right internal iliac vessels. Endoscopic ultrasound and magnetic resonance imaging excluded any involvement of the pelvic colon or residual rectum. A combined extraperitoneal antero-lateral approach and gluteal access were used to optimize vascular control on the internal iliac vessels, to promptly identify the ureter and to achieve a better posterior exposition of the sciatic notch. This technique allowed a controlled and tailored resection of pelvic sidewall without entering into the abdominal cavity. The postoperative course was uneventful. The pathologic report confirmed clear margins (R0), with one involving obturator lymph node. At 3months, the patient is alive and free from local re-relapse. A right lung metastasis has occurred, and it was treated by stereotactic radiotherapy. The present report proposes a novel extraperitoneal pelvic sidewall excision to resect lateral LRRC with a colorectal-sparing approach, thus minimizing the risk of exenterative surgery-related complications. A proper selection of patients is mandatory, as the proposed technique could not be generalized as the standard of care in all lateral LRRCs.

Feasibility and clinical value of linear endoscopic ultrasonography imaging in the lower gastrointestinal subepithelial lesions.

Linear endoscopic ultrasonography (EUS) has been extensively utilized as a novel diagnostic and therapeutic modality across various fields. However, there have been relatively few studies focusing on lower gastrointestinal lesions. The aim of our study was to investigate the feasibility, safety and clinical value of linear EUS in the lower gastrointestinal subepithelial lesions. This was a retrospective study involving patients with lower gastrointestinal subepithelial lesions diagnosed by linear EUS from August 2019 to April 2023 at the Second Affiliated Hospital of Anhui Medical University. The data, including basic clinical information, linear EUS features, technical success rate, complications, and follow-up, were retrospectively collected and analyzed. A total of 69 patients with lower gastrointestinal subepithelial lesions underwent examination by linear EUS. Excluding the rectum, the technical success rate of linear EUS was 90.6% (29/32). Apart from the 7 patients whose diagnosis remained unknown, 3 patients with no abnormal EUS findings, and 3 patients failed the procedure, 56 patients were included in the final diagnostic performance analysis. The most common locations of the lesions were the rectum (37/56, 66.1%) and sigmoid colon (7/56, 12.5%). Based on endoscopy findings and pathological results, the most prevalent types of subepithelial lesions in the lower gastrointestinal tract were neuroendocrine tumor (NET) (12/56, 20.3%), lipoma (8/56, 13.6%) and extraluminal compression (8/56, 13.6%). The majority of lesions ranged in diameter from 1 to 3cm (χ2 = 18.750, p < 0.001). After undergoing linear EUS examination, 36 patients received EUS-FNA (3/36), biopsy (5/36), endoscopic resection (25/36), or surgical excision (3/36) respectively. The pathological results of 29 patients were entirely consistent with the diagnosis made using linear EUS, with an 80.6% (29/36) diagnostic accuracy rate. Follow-up indicated that the lesions remained unchanged within 6-36months. All patients tolerated the procedure well without any complications. In conclusion, linear EUS demonstrates technical feasibility, safety, and a high diagnostic accuracy for subepithelial lesions in the lower gastrointestinal tract.

Differentiating localized autoimmune pancreatitis and pancreatic ductal adenocarcinoma using endoscopic ultrasound images with deep learning.

Localized autoimmune pancreatitis is difficult to differentiate from pancreatic ductal adenocarcinoma on endoscopic ultrasound images. In recent years, deep learning methods have improved the diagnosis of diseases. Hence, we developed a special cross-validation framework to search for effective methodologies of deep learning in distinguishing autoimmune pancreatitis from pancreatic ductal adenocarcinoma on endoscopic ultrasound images. Data from 24 patients diagnosed with localized autoimmune pancreatitis(8751 images) and 61 patients diagnosed with pancreatic ductal adenocarcinoma (20,584 images) were collected from 2016 to 2022. We applied transfer learning to a convolutional neural network called ResNet152, together with our innovative imaging method contributing to data augmentation and temporal data process. We divided patients into five groups according to different factors for 5-fold cross-validation, where the ordered and balanced datasets were created for the performance evaluations. ResNet152 surpassed the endoscopists in all evaluation metrics with almost all datasets. Interestingly, when the dataset is balanced according to the factor of the endoscopists' diagnostic accuracy, the area under the receiver operating characteristic curve and accuracy were highest at 0.85 and 0.80, respectively. It is deduced that image features useful for ResNet152 correlate with those used by endoscopists for their diagnoses. This finding may contribute to sample-efficient dataset preparation to train convolutional neural networks for endoscopic ultrasonography-imaging diagnosis.

A311 EFFICACY OF PANCREATIC CANCER SCREENING IN HIGH RISK INDIVIDUALS

Abstract Background Pancreatic cancer (PC) is the 11th most prevalent cancer in Canada with about 6900 new cases in 2021. Due to late diagnosis, it is the 4th largest cause of cancer related mortality in Canada. Despite advances in diagnosis and treatment over the past decade, the 5-year survival remains ampersand:003C10%, making pancreatic cancer one of the most lethal malignancies worldwide. However, if detected early, the 5-year survival increases to 80%. The incidence of pancreatic cancer remains low. Therefore, the International Cancer of the Pancreas Screening (CAPS) consortium do not recommend screening everyone. Selective screening in high-risk individuals may be effective in early detection and treatment of pancreatic cancer. Aims 1. To determine the characteristics of patients undergoing PC screening at a single tertiary care center in Vancouver. 2. To determine the frequency of disease progression and outcomes in patients diagnosed with high-risk lesions during screening. Methods Retrospective chart review was performed for patients who were screened between January 1, 2005 to June 30, 2023. As per consensus from the CAPS Consortium, patients with at least 2 close relatives with PC, and those with predisposing genetic mutations screened using endoscopic ultrasound (EUS) and/or magnetic resonance imaging (MRI) were included in the study. Clinical data, such as age, tobacco use, genetic risk, endoscopic findings, and outcomes, such as disease progression, were collected. Statistical analysis was performed using Rstudio and Microsoft Excel. Results 73 patients were included, with 42 included due to pathogenic germline mutations, and 31 with familial pancreatic cancer. Most were female (52, 71.2%). Median age in the study was 54 years (range 24-74 years) at initial screening. 66 patients (90.4%) received an EUS and 7 patients (9.6%) received MRI for initial screening. At the time of initial screening. 74.6% of patients had a normal pancreas. 6 patients (8.2%) were found to have neoplastic lesions, e.g. ampulloma, neuroendocrine tumour or adenocarcinoma at initial screening and referred for surgery. In the median follow up time of 67.9 months (range 0-168.1 months), 6 patients were noted to have disease progression, 4 of which had increase in cyst size without any other worrisome features. 2 patients had progression to malignancy. 1 of those patients received chemotherapy and surgical resection and remains alive; the other patient was lost of follow up due to geographical move. Conclusions Pancreatic cancer screening in high risk individuals is effective in early detection and treatment. Given lower incidence of pancreatic cancer, we likely need additional patient enrolment to observe a significant improvement in patient outcomes. Funding Agencies None

A fast regional convolutional neural network system for the diagnosis of gastrointestinal subepithelial lesions on endoscopic ultrasound images.

A fast regional convolutional neural network system for the diagnosis of gastrointestinal subepithelial lesions on endoscopic ultrasound images.

DSMT-Net: Dual Self-Supervised Multi-Operator Transformation for Multi-Source Endoscopic Ultrasound Diagnosis.

Pancreatic cancer has the worst prognosis of all cancers. The clinical application of endoscopic ultrasound (EUS) for the assessment of pancreatic cancer risk and of deep learning for the classification of EUS images have been hindered by inter-grader variability and labeling capability. One of the key reasons for these difficulties is that EUS images are obtained from multiple sources with varying resolutions, effective regions, and interference signals, making the distribution of the data highly variable and negatively impacting the performance of deep learning models. Additionally, manual labeling of images is time-consuming and requires significant effort, leading to the desire to effectively utilize a large amount of unlabeled data for network training. To address these challenges, this study proposes the Dual Self-supervised Multi-Operator Transformation Network (DSMT-Net) for multi-source EUS diagnosis. The DSMT-Net includes a multi-operator transformation approach to standardize the extraction of regions of interest in EUS images and eliminate irrelevant pixels. Furthermore, a transformer-based dual self-supervised network is designed to integrate unlabeled EUS images for pre-training the representation model, which can be transferred to supervised tasks such as classification, detection, and segmentation. A large-scale EUS-based pancreas image dataset (LEPset) has been collected, including 3,500 pathologically proven labeled EUS images (from pancreatic and non-pancreatic cancers) and 8,000 unlabeled EUS images for model development. The self-supervised method has also been applied to breast cancer diagnosis and was compared to state-of-the-art deep learning models on both datasets. The results demonstrate that the DSMT-Net significantly improves the accuracy of pancreatic and breast cancer diagnosis.

Comparative Analysis of Baseline Vnet and Unet Architectures on Pancreas Segmentation

The pancreas is one of the vital organs in the human body. It has an essential role in the digestive system and endocrine system. Diseases such as cancer, diabetes, hormonal problems, pancreatitis, and digestive problems occur in pancreatic disorders. In detecting pancreatic disorders, first blood and urine tests are requested. If further examination is needed, CT (Computed Tomography), MR (Magnetic Resonance), and EUS (Endoscopic Ultrasonography) imaging methods are used. Pancreas segmentation is generally the process of defining and drawing the lines of the pancreas from medical images such as CT and MRI. The size and shape of the pancreas varies from person to person. Manual segmentation of the pancreas is time-consuming and varies between physicians. Recently, deep learning-based segmentation methods that achieve high-performance results in organ segmentation have become trendy. In this study, Unet and Vnet architectures were comparatively analyzed on the NIH-CT-82 dataset. As a result of the ablation studies, a validation sensitivity of 0.9978 and a validation loss of 0.041 were obtained in the Unet architecture. In the training with the Vnet architecture, 0.9975 validation sensitivity and 0.046 validation loss values were obtained, respectively.

Enhancing gastrointestinal submucosal tumor recognition in endoscopic ultrasonography: A novel multi-attribute guided contextual attention network

Endoscopic ultrasonography (EUS) is a valuable imaging modality for diagnosing gastrointestinal submucosal tumors (SMTs). However, inherent content variations in EUS images due to gastrointestinal tract mobility and handheld ultrasound instability challenge the saliency of SMTs’ visual features. The presence of fine-grained inter-class and large intra-class differences further complicates EUS-based diagnosis. To address these issues, this paper presents a novel Multi-Attribute Guided Contextual Attention Network (MAG-CA-Net) for interpretable SMT recognition in EUS. Inspired by endoscopists’ clinical diagnosis expertise, our framework initially localizes abnormal areas based on echo attributes and subsequently determines tumor categories using contextual semantics. Experimental results demonstrate the effectiveness of MAG-CA-Net, exhibiting improved recognition recall and precision rates for gastrointestinal stromal tumor, leiomyoma, and pancreatic rest. Specifically, the MAG network facilitates abnormal area localization, while the CA mechanism guides the model to focus on the most discriminative tumor-context-associated regions. The proposed method achieved an average classification accuracy of 93.16%, an average precision of 93.17%, a weighted recall of 93.16%, and an average F1-score of 93.15 % for the three disease categories. The proposed approach provides crucial guidelines for data collection standards and model development in the clinical diagnosis process of SMTs under EUS. Its interpretability analysis enhances the credibility of clinical physicians towards assisted diagnostic methods based on deep learning. The source code will be publicly available at https://gitee.com/HangbinZheng/mag-ca-net.

Approach to the Patient: Insulinoma.

Insulinomas are hormone-producing pancreatic neuroendocrine neoplasms with an estimated incidence of 1 to 4 cases per million per year. Extrapancreatic insulinomas are extremely rare. Most insulinomas present with the Whipple triad: (1) symptoms, signs, or both consistent with hypoglycemia; (2) a low plasma glucose measured at the time of the symptoms and signs; and (3) relief of symptoms and signs when the glucose is raised to normal. Nonmetastatic insulinomas are nowadays referred to as "indolent" and metastatic insulinomas as "aggressive." The 5-year survival of patients with an indolent insulinoma has been reported to be 94% to 100%; for patients with an aggressive insulinoma, this amounts to 24% to 67%. Five percent to 10% of insulinomas are associated with the multiple endocrine neoplasia type 1 syndrome. Localization of the insulinoma and exclusion or confirmation of metastatic disease by computed tomography is followed by endoscopic ultrasound or magnetic resonance imaging for indolent, localized insulinomas. Glucagon-like peptide 1 receptor positron emission tomography/computed tomography or positron emission tomography/magnetic resonance imaging is a highly sensitive localization technique for seemingly occult, indolent, localized insulinomas. Supportive measures and somatostatin receptor ligands can be used for to control hypoglycemia. For single solitary insulinomas, curative surgical excision remains the treatment of choice. In aggressive malignant cases, debulking procedures, somatostatin receptor ligands, peptide receptor radionuclide therapy, everolimus, sunitinib, and cytotoxic chemotherapy can be valuable options.

Endoscopic Ultrasonography-Guided Fine-Needle Aspiration for Duodenal Subepithelial Lesions Showing a Hypoechoic Mass on Endoscopic Ultrasound Imaging.

For duodenal subepithelial lesions showing a hypoechoic mass on endoscopic ultrasound imaging, the utility of endoscopic ultrasound-guided fine-needle aspiration and the frequency of histological types have not been the focus of previous literature. This study aimed to clarify this. This prospective observational study enrolled 22 consecutive patients who underwent endoscopic ultrasoundguided fine-needle aspiration for duodenal subepithelial lesions with hypoechoic mass on endoscopic ultrasound. Immunohistochemical analysis was performed for all endoscopic ultrasound-guided fine-needle aspiration and surgically resected specimens. The main outcome measures were the technical results of endoscopic ultrasound-guided fine-needle aspiration and the frequency of histological types of duodenal subepithelial lesions with hypoechoic mass. Thirteen fine-needle aspiration specimens were obtained from the duodenal bulb and eight from the descending duodenal region. The puncture was not performed because of intervening vessels in one patient. The diagnostic rate was 81% (95% confidence interval: 58.1-94.6, 17/21 patients). In 12 patients receiving surgical resection (excluding one cancellation of endoscopic ultrasoundguided fine-needle aspiration), the diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration was 75% (95% confidence interval: 42.8-94.5, 9/12 patients). No complications were observed. The histopathological diagnoses included 11 cases of gastrointestinal stromal tumor (50%), 2 cases of leiomyoma (9%), 2 cases of metastatic cancer (9%), 2 cases of benign inconclusive, and 1 case each of carcinoid, malignant lymphoma, leiomyosarcoma, gauzeoma, and aberrant pancreas (4.5% each). The frequency of malignant tumors in the duodenal subepithelial lesions with hypoechoic mass group was 73% (16/22 patients). Endoscopic ultrasound-guided fine-needle aspiration for duodenal subepithelial lesions with hypoechoic mass was safe and accurate. As duodenal subepithelial lesion with hypoechoic mass has a reasonably high possibility of containing malignant tumors, it is desirable to perform endoscopic ultrasound-guided fine-needle aspiration.

Experiences in endoscopic ultrasonography at a tertiary center general surgery endoscopy unit

Aim: Endoscopic ultrasonography (EUS) is used for diagnosing pathologies in the gastrointestinal system and surrounding organs. This study aims to share the experiences and results of EUS and biopsy at a tertiary center general surgery endoscopy unit, in the context of the literature.&#x0D; Methods: Patients who underwent EUS imaging at the general surgery endoscopy unit between January 2021 and January 2022 were retrospectively reviewed. Demographic characteristics, EUS biopsy indications, clinical pre-diagnoses, preoperative imaging methods, biopsy counts, and results, as well as complications, were analyzed.&#x0D; Results: The mean age of the 292 patients was 56.5 years, with 53.8% being male. The most common EUS indication was a mass in the pancreas, accounting for 62%. Endoscopic fine-needle aspiration biopsy was performed on 127 (43.4%) patients, and the mean diameter of biopsied masses was 35.1 mm. Malignancy was detected in 58.2% of the biopsied patients. Based on biopsy and lesion characteristics, 49.3% of patients were followed up, 27.4% underwent surgery, 15.4% received oncological treatment, and 7.9% were given endoscopic retrograde cholangiopancreatography.&#x0D; Conclusion: The role of EUS in the diagnosis and treatment of gastrointestinal malignancies, pancreatic diseases, and biliary diseases continues to evolve. We believe that EUS plays a key role in the multidisciplinary management of complex surgical and oncology patients and those with pancreatobiliary disorders.

Application of robotic (or laparoscopic) surgery combined with colonoscopy in T1 stage colorectal cancer surgery: 13 cases

Objective: To investigate the feasibility and safety of a robotic surgical system (or laparoscopy) in combination with colonoscopy (combined) for the treatment of stage T1N0M0 colorectal cancer. Methods: This was a descriptive case series. Indications for combined dual-scope surgery in this study were as follows: (1) preoperative colonoscopic examination of lesions in the middle and upper rectum and colon with pathologically confirmed high-grade intraepithelial neoplasia, intramucosal adenocarcinoma, or adenocarcinoma; (2) no distant or local lymph node metastases; and (3) endoscopic ultrasound and magnetic resonance imaging evidence of tumor invasion of the mucosal or submucosal, but not the muscular, layer (i.e., T1). The clinical data of 13 patients with stage T1 colorectal cancer who had undergone dual-scope combined resection using a robotic surgery system or laparoscope-assisted combined colonoscopy surgery at the First Affiliated Hospital of Zhengzhou University from April to October 2022 were retrospectively collected, including 6 males and 7 females, with a median age of 59 (48~88) years old. The tumors were located in the upper and middle rectum in six patients, in the sigmoid colon in three, and in the ascending colon in four. The median maximum diameter of the tumors was 3.0 (1.8-5.0) cm. The surgery was performed by a robotic surgery system (or laparoscopy) with peritumoral D1 lymph node dissection at the first station in the tumor area. The tumors were resected under direct vision and the defects in the intestinal wall were using a robotic surgery system (or laparoscopy). A robotic surgery system was combined with colonoscopy in eight cases and laparoscopy combined with colonoscopy in the remaining five. Studied variables includes surgical and pathological features, postoperative factors, and outcomes. Results: Surgery was successful in all 13 patients with no need for conversion to open surgery or intraoperative blood transfusion. The median operating time was 85 (60-120) minutes, median intraoperative bleeding 3 (2-5) mL, median number of lymph nodes harvested 3 (1-5), and the median circumferential resection margin 0.8 (0.5-1.0) cm. Postoperative pathological examination showed lymph node metastasis in one patient, who therefore underwent additional radical surgery. The median postoperative time to ambulation was 1 (1-2) days. The urinary catheters of all patients were removed 1 day after surgery and the median length of stay was 4 (3-5) days. No abdominal infection, anastomotic leakage or bleeding occurred in any of the study patients. The median follow-up time was 10 (6-12) months, during which no tumor recurrence or metastasis was found, and the quality of life was satisfactory. Conclusions: The combination of two minimally invasive platforms, a robotic surgery system (or laparoscopy) and colonoscopy, is safe and feasible for resection of stage T1 colorectal cancer and has a good short-term prognosis.

Differentiating Gastrointestinal Stromal Tumors From Leiomyomas of Upper Digestive Tract Using Convolutional Neural Network Model by Endoscopic Ultrasonography.

Gastrointestinal stromal tumors (GISTs) and leiomyomas are the most common submucosal tumors of the upper digestive tract, and the diagnosis of the tumors is essential for their treatment and prognosis. However, the ability of endoscopic ultrasonography (EUS) which could correctly identify the tumor types is limited and closely related to the knowledge, operational level, and experience of the endoscopists. Therefore, the convolutional neural network (CNN) is used to assist endoscopists in determining GISTs or leiomyomas with EUS. A model based on CNN was constructed according to GoogLeNet architecture to distinguish GISTs or leiomyomas. All EUS images collected from this study were randomly sampled and divided into training set (n=411) and testing set (n=103) in a ratio of 4:1. The CNN model was trained by EUS images from the training set, and the testing set was utilized to evaluate the performance of the CNN model. In addition, there were some comparisons between endoscopists and CNN models. It was shown that the sensitivity and specificity in identifying leiomyoma were 95.92%, 94.44%, sensitivity and specificity in identifying GIST were 94.44%, 95.92%, and accuracy in total was 95.15% of the CNN model. It indicates that the diagnostic accuracy of the CNN model is equivalent to skilled endoscopists, or even higher than them. While identifying GIST or leiomyoma, the performance of CNN model was robust, which is highlighting its promising role in supporting less-experienced endoscopists and reducing interobserver agreement.

Proposal for T3 classification of esophagogastric junction carcinoma based on the interconnection of extramural anatomical structures.

Classification of extramural invasion of esophagogastric junction carcinoma (EGJC) is not yet established. The anatomy surrounding the EGJ alters between the mediastinum and the abdominal cavity. This review proposed a T3 classification of EGJC based on anatomical continuity. Analysis of endoscopic ultrasound images, review of intraoperative images, and detailed observation of surgical specimens were followed by a review of the literature. In the EGJ, the muscularis propria of the esophagus is enclosed in mediastinal adipose tissue called the adventitia, which is surrounded by the diaphragmatic crus and contains the paraesophageal lymph nodes (LNs). After passing through the esophageal hiatus along with the vagus nerves and blood vessels, the adventitia joins the adipose tissue containing the paracardial LNs, which is covered by the peritoneum, and then further divides into the lesser and greater omentum. The connective tissue outside the muscularis propria of the stomach, including the adipose tissue of the omentum, is called the subserosa. According to the TNM classification, T3 esophageal and gastric cancer is defined as invasion of the adventitia and subserosa, respectively. Given that the adventitia is anatomically continuous with the subserosa, T3 tumors of the EGJ can be described as those that extend through the muscularis propria but do not invade the peritoneum or diaphragmatic crus. We propose classifying T3 EGJC as "tumor extends through muscularis propria" rather than using the separate terms "adventitia" and "submucosa". T4 could be "tumor perforates serosa or invades adjacent structures", as per the current gastric cancer classification.

A Comprehensive Guide to Artificial Intelligence in Endoscopic Ultrasound.

Endoscopic Ultrasound (EUS) is widely used for the diagnosis of bilio-pancreatic and gastrointestinal (GI) tract diseases, for the evaluation of subepithelial lesions, and for sampling of lymph nodes and solid masses located next to the GI tract. The role of Artificial Intelligence in healthcare in growing. This review aimed to provide an overview of the current state of AI in EUS from imaging to pathological diagnosis and training. AI algorithms can assist in lesion detection and characterization in EUS by analyzing EUS images and identifying suspicious areas that may require further clinical evaluation or biopsy sampling. Deep learning techniques, such as convolutional neural networks (CNNs), have shown great potential for tumor identification and subepithelial lesion (SEL) evaluation by extracting important features from EUS images and using them to classify or segment the images. AI models with new features can increase the accuracy of diagnoses, provide faster diagnoses, identify subtle differences in disease presentation that may be missed by human eyes, and provide more information and insights into disease pathology. The integration of AI in EUS images and biopsies has the potential to improve the diagnostic accuracy, leading to better patient outcomes and to a reduction in repeated procedures in case of non-diagnostic biopsies.