Dataset Information

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

ABSTRACT:

Background and aims

Distinguishing pancreatic cancer from nonneoplastic masses is critical and remains a clinical challenge. The study aims to construct a deep learning-based artificial intelligence system to facilitate pancreatic mass diagnosis, and to guide EUS-guided fine-needle aspiration (EUS-FNA) in real time.

Methods

This is a prospective study. The CH-EUS MASTER system is composed of Model 1 (real-time capture and segmentation) and Model 2 (benign and malignant identification). It was developed using deep convolutional neural networks and Random Forest algorithm. Patients with pancreatic masses undergoing CH-EUS examinations followed by EUS-FNA were recruited. All patients underwent CH-EUS and were diagnosed both by endoscopists and CH-EUS MASTER. After diagnosis, they were randomly assigned to undergo EUS-FNA with or without CH-EUS MASTER guidance.

Results

Compared with manual labeling by experts, the average overlap rate of Model 1 was 0.708. In the independent CH-EUS video testing set, Model 2 generated an accuracy of 88.9% in identifying malignant tumors. In clinical trial, the accuracy, sensitivity, and specificity for diagnosing pancreatic masses by CH-EUS MASTER were significantly better than that of endoscopists. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were respectively 93.8%, 90.9%, 100%, 100%, and 83.3% by CH-EUS MASTER guided EUS-FNA, and were not significantly different compared to the control group. CH-EUS MASTER-guided EUS-FNA significantly improved the first-pass diagnostic yield.

Conclusion

CH-EUS MASTER is a promising artificial intelligence system diagnosing malignant and benign pancreatic masses and may guide FNA in real time.

Trial registration number

NCT04607720.

SUBMITTER: Tang A

PROVIDER: S-EPMC10134340 | biostudies-literature | 2023 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

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

Tang Anliu A Tian Li L Gao Kui K Liu Rui R Hu Shan S Liu Jinzhu J Xu Jiahao J Fu Tian T Zhang Zinan Z Wang Wujun W Zeng Long L Qu Weiming W Dai Yong Y Hou Ruirui R Tang Shoujiang S Wang Xiaoyan X

Cancer medicine 20230106 7

<h4>Background and aims</h4>Distinguishing pancreatic cancer from nonneoplastic masses is critical and remains a clinical challenge. The study aims to construct a deep learning-based artificial intelligence system to facilitate pancreatic mass diagnosis, and to guide EUS-guided fine-needle aspiration (EUS-FNA) in real time.<h4>Methods</h4>This is a prospective study. The CH-EUS MASTER system is composed of Model 1 (real-time capture and segmentation) and Model 2 (benign and malignant identificat ...[more]

PMID: 36606571

Similar Datasets

Project description:Long term prognosis and 5-year survival for pancreatic adenocarcinoma (PDAC) remains suboptimal. Endoscopic ultrasound (EUS) guided RFA (EUS-RFA) is an emerging technology and limited data exist regarding safety and long-term outcomes. The aim of this study is to report safety-profile, feasibility and outcomes of EUS-RFA for advanced PDAC. Prospective review of patients with diagnosis of locally-advanced or metastatic PDAC undergoing EUS-RFA between October 2016 to March 2018 with long-term follow up (> 30 months). Study patients underwent a total of 1-4 RFA sessions. All patients were enrolled in longitudinal cohort study and received standard of care chemotherapy. 10 patients underwent EUS-RFA. Location of the lesions was in the head(4), neck(2), body(2), and tail(2). 22 RFA sessions were performed with a range of 1-4 sessions per patient. There were no major adverse events (bleeding, perforation, infection, pancreatitis) in immediate (up to 72 h) and short-term follow up (4 weeks). Mild worsening of existing abdominal pain was noted during post-procedure observation in 12/22 (55%) of RFA treatments. Follow-up imaging demonstrated tumor progression in 2 patients, whereas tumor regression was noted in 6 patients (> 50% reduction in size in 3 patients). Median survival for the cohort was 20.5 months (95% CI, 9.93-42.2 months). Currently, 2 patients remain alive at 61 and 81 months follow-up since initial diagnosis. One patient had 3 cm PDAC with encasement of the portal confluence, abutment of the celiac axis, common hepatic and superior mesenteric artery. This patient had significant reduction in tumor size and underwent standard pancreaticoduodenectomy. In our experience, EUS-RFA was safe, well-tolerated and could be concurrently performed with standard chemotherapy. In this select cohort, median survival was improved when compared to published survival based upon SEER database and clinical trials. Future prospective trials are needed to understand the role of EUS-RFA in overall management of PDAC.

Dataset Information

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

Background and aims

Methods

Results

Conclusion

Trial registration number

Publications

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets