Accurate prediction of histological grading of intraductal papillary mucinous neoplasia using deep learning

Schulz, Dominik; Heilmaier, Markus; Phillip, Veit; Treiber, Matthias; Mayr, Ulrich; Lahmer, Tobias; Müeller, Jens-Dominik; Demir, Ihsan Ekin; Frieß, Helmut; Reichert, Maximilian; Schmid, Roland M.; Abdelhafez, Mohamed

doi:10.1055/a-1971-1274

Cited by 10 publications

(6 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In a study of 43 patients with IPMNs who underwent pancreatectomy, a model was trained using 3355 EUS images from these patients, resulting in the convolutional neural network classifying high-grade dysplasia with an accuracy of 99.6%. The AI model was superior in risk stratifying IPMNs than the guidelines, with accuracies ranging between 51.8% and 70.3% [ 32 ]. In a separate investigation involving 50 IPMNs and 3790 EUS images, a convolutional neural network model designed for IPMN risk assessment demonstrated impressive performance metrics.…”

Section: Applications Of Artificial Intelligence For Eus and Clementioning

confidence: 99%

Endoscopic Ultrasound-Guided Needle-Based Confocal Endomicroscopy as a Diagnostic Imaging Biomarker for Intraductal Papillary Mucinous Neoplasms

Krishna,

Abdelbaki,

Hart

et al. 2024

Cancers

View full text Add to dashboard Cite

Pancreatic cancer is on track to become the second leading cause of cancer-related deaths by 2030, yet there is a lack of accurate diagnostic tests for early detection. Intraductal papillary mucinous neoplasms (IPMNs) are precursors to pancreatic cancer and are increasingly being detected. Despite the development and refinement of multiple guidelines, diagnosing high-grade dysplasia or cancer in IPMNs using clinical, radiologic, endosonographic, and cyst fluid features still falls short in terms of accuracy, leading to both under- and overtreatment. EUS-guided needle-based confocal laser endomicroscopy (nCLE) is a novel technology that allows real-time optical biopsies of pancreatic cystic lesions. Emerging data has demonstrated that EUS-nCLE can diagnose and risk stratify IPMNs more accurately than conventional diagnostic tools. Implementing EUS-nCLE in clinical practice can potentially improve early diagnosis of pancreatic cancer, reduce unnecessary surgeries of IPMNs with low-grade dysplasia, and advance the field of digital pathomics. In this review, we summarize the current evidence that supports using EUS-nCLE as a diagnostic imaging biomarker for diagnosing IPMNs and for risk stratifying their degree of neoplasia. Moreover, we will present emerging data on the role of adding artificial intelligence (AI) algorithms to nCLE and integrating novel fluid biomarkers into nCLE.

show abstract

Section: Applications Of Artificial Intelligence For Eus and Clementioning

confidence: 99%

Endoscopic Ultrasound-Guided Needle-Based Confocal Endomicroscopy as a Diagnostic Imaging Biomarker for Intraductal Papillary Mucinous Neoplasms

Krishna,

Abdelbaki,

Hart

et al. 2024

Cancers

View full text Add to dashboard Cite

show abstract

“…For the differential diagnosis between benign and malignant IPMN using EUS images, the usefulness of deep learning for overcoming poor objectivity has already been reported. 59 , 60 Schulz et al. performed machine learning using 3355 EUS images from 43 patients.…”

Section: Future Perspectivesmentioning

confidence: 99%

Current status of endoscopic ultrasound in the diagnosis of intraductal papillary mucinous neoplasms

Ohno,

Kuzuya,

Kawabe

et al. 2024

DEN Open

View full text Add to dashboard Cite

The new Kyoto guidelines for the management of intraductal papillary mucinous neoplasm (IPMN) provide evidence‐based recommendations for the diagnosis and treatment of IPMN. Endoscopic ultrasonography (EUS) is a diagnostic modality with a high spatial resolution that allows detailed observation and obtaining cyst fluid or tissue samples via EUS‐guided fine needle aspiration (EUS‐FNA). Currently, EUS is an indispensable examination method for the diagnosis of pancreatic diseases. On the other hand, there have been concerns that EUS imaging tends to be highly operator‐dependent, and may lack objectivity. Previous guidelines have assigned EUS as an option for patients with worrisome features. However, recent reports indicate that the sensitivity of EUS for the diagnosis of mural nodules (MNs) is more than 90%, comparable or superior to that of contrast‐enhanced computed tomography or magnetic resonance cholangiopancreatography. The specific advantages of EUS in the diagnosis of IPMN are: (1) high spatial resolution imaging for the diagnosis of MNs, (2) contrast‐enhanced EUS for differentiation of intra‐cystic MNs from mucous clots, and (3) pathological diagnosis using EUS‐FNA and differential diagnosis of a pancreatic cystic tumor by cystic fluid analysis. In order to utilize EUS in the diagnosis of IPMN, endoscopists are required to have the skills to provide sufficiently objective imaging findings.

show abstract

“…Most research until today were conducted on CT images, with a few investigations involving MRI [33 ▪▪ ] and EUS [27,28]. Several studies have developed end-to-end DL solutions for pancreatic tumor detection and classification.…”

Section: Deep Learningmentioning

confidence: 99%

A review of deep learning and radiomics approaches for pancreatic cancer diagnosis from medical imaging

Zhang

Keleş

et al. 2023

Current Opinion in Gastroenterology

View full text Add to dashboard Cite

Purpose of review Early and accurate diagnosis of pancreatic cancer is crucial for improving patient outcomes, and artificial intelligence (AI) algorithms have the potential to play a vital role in computer-aided diagnosis of pancreatic cancer. In this review, we aim to provide the latest and relevant advances in AI, specifically deep learning (DL) and radiomics approaches, for pancreatic cancer diagnosis using cross-sectional imaging examinations such as computed tomography (CT) and magnetic resonance imaging (MRI). Recent findings This review highlights the recent developments in DL techniques applied to medical imaging, including convolutional neural networks (CNNs), transformer-based models, and novel deep learning architectures that focus on multitype pancreatic lesions, multiorgan and multitumor segmentation, as well as incorporating auxiliary information. We also discuss advancements in radiomics, such as improved imaging feature extraction, optimized machine learning classifiers and integration with clinical data. Furthermore, we explore implementing AI-based clinical decision support systems for pancreatic cancer diagnosis using medical imaging in practical settings. Summary Deep learning and radiomics with medical imaging have demonstrated strong potential to improve diagnostic accuracy of pancreatic cancer, facilitate personalized treatment planning, and identify prognostic and predictive biomarkers. However, challenges remain in translating research findings into clinical practice. More studies are required focusing on refining these methods, addressing significant limitations, and developing integrative approaches for data analysis to further advance the field of pancreatic cancer diagnosis.

show abstract

Accurate prediction of histological grading of intraductal papillary mucinous neoplasia using deep learning

Cited by 10 publications

References 35 publications

Endoscopic Ultrasound-Guided Needle-Based Confocal Endomicroscopy as a Diagnostic Imaging Biomarker for Intraductal Papillary Mucinous Neoplasms

Endoscopic Ultrasound-Guided Needle-Based Confocal Endomicroscopy as a Diagnostic Imaging Biomarker for Intraductal Papillary Mucinous Neoplasms

Current status of endoscopic ultrasound in the diagnosis of intraductal papillary mucinous neoplasms

A review of deep learning and radiomics approaches for pancreatic cancer diagnosis from medical imaging

Contact Info

Product

Resources

About