Publications by authors named "Dana M Middendorf"
Gastrointest Endosc
July 2021
Article Synopsis
- EUS-guided needle-based confocal laser endomicroscopy (EUS-nCLE) can help identify high-grade dysplasia/adenocarcinoma (HGD-Ca) in intraductal papillary mucinous neoplasms (IPMNs), but traditionally requires manual analysis.
- Researchers developed two computer-aided diagnosis (CAD) algorithms using convolutional neural networks (CNNs) to automate and improve the accuracy of IPMN diagnosis and risk assessment.
- The study showed that these CNN-CAD algorithms outperformed established guidelines in sensitivity and accuracy for detecting HGD-Ca, indicating their potential for real-time clinical application with future enhancements.
