Radiomics in stratification of pancreatic cystic lesions: Machine learning in action

Abstract

Pancreatic cystic lesions (PCLs) are well-known precursors of pancreatic cancer. Their diagnosis can be challenging as their behavior varies from benign to malignant disease. Precise and timely management of malignant pancreatic cysts might prevent transformation to pancreatic cancer. However, the current consensus guidelines, which rely on standard imaging features to predict cyst malignancy potential, are conflicting and unclear. This has led to an increased interest in radiomics, a high-throughput extraction of comprehensible data from standard-of-care images. Radiomics can be used as a diagnostic and prognostic tool in personalized medicine. It utilizes quantitative image analysis to extract features in conjunction with machine learning and artificial intelligence (AI) methods like support vector machines, random forest, and convolutional neural networks for feature selection and classification. Selected features can then serve as imaging

Artificial intelligence in pancreaticobiliary endoscopy

“…EUS fine‐needle aspiration and biopsy specimens are often nondiagnostic, and AI‐based methods were applied to cytological or histological specimen images to improve the diagnostic yield or automatize histological characterization 23,24 . Radiomics again has been at the forefront with improved characterization, classification, and prognostication of pancreatic cystic lesions 25 . A recent study used CNN on 3970 EUS images to identify high‐risk lesions within intraductal papillary mucinous neoplasms 26 .…”

“…23,24 Radiomics again has been at the forefront with improved characterization, classification, and prognostication of pancreatic cystic lesions. 25 A recent study used CNN on 3970 EUS images to identify high-risk lesions within intraductal papillary mucinous neoplasms. 26 The authors report an area under the receiver operating characteristic curve of 0.98 and that AI-based diagnosis had higher accuracy than human diagnosis (56.0%).…”

Künstliche Intelligenz in der Gastroenterologie

“…Hier kann KI menschliche Befundung automatisieren und objektivieren [6]. Ein wichtiges Beispiel stellen zystische Pankreasläsionen dar, welche in maligne Pankreaskarzinome übergehen können und deren genaue Zuordnung anhand bildmorphologischer Kriterien oft schwierig ist [7]. Weitere Studien konnten zeigen, dass Leberläsionen mittels KI automatisiert detektiert und klarer eingegrenzt werden können [8].…”

Symptom Relief is the Most Significant Prognostic Factor for Unresectable Locally Advanced/Recurrent Pancreatic Cancer Receiving Chemoradiotherapy: Analysis of 65 Cases

“…Vipin Dalal et al reviewed studies using radiomics to predict the short-term e cacy, local control, DFS and OS of pancreatic cancer patients with different stages to antitumor therapy. [12] Allen Li's team used daily CT image extraction parameters to predict tumor withdrawal after CRT, [13] which we can consider to apply in further study, to stratify patients before treatment and try to optimize treatment in clinical studies.…”