Artificial Intelligence in the Diagnosis and Treatment of Pancreatic Cystic Lesions and Adenocarcinoma

Abstract

Pancreatic cancer is projected to become the second leading cause of cancer-related mortality in the United States by 2030. This is in part due to the paucity of reliable screening and diagnostic options for early detection. Amongst known pre-malignant pancreatic lesions, pancreatic intraepithelial neoplasia (PanIN) and intraductal papillary mucinous neoplasms (IPMNs) are the most prevalent. The current standard of care for the diagnosis and classification of pancreatic cystic lesions (PCLs) involves cross-sectional imaging studies and endoscopic ultrasound (EUS) and, when indicated, EUS-guided fine needle aspiration and cyst fluid analysis. However, this is suboptimal for the identification and risk stratification of PCLs, with accuracy of only 65–75% for detecting mucinous PCLs. Artificial intelligence (AI) is a promising tool that has been applied to improve accuracy in screening for solid tumors, including breast, lung, cervical, and colon cancer. More recently, it has shown promise in diagnosing pancreatic cancer by identifying high-risk populations, risk-stratifying premalignant lesions, and predicting the progression of IPMNs to adenocarcinoma. This review summarizes the available literature on artificial intelligence in the screening and prognostication of precancerous lesions in the pancreas, and streamlining the diagnosis of pancreatic cancer.

Molecular Pathology of Pancreatic Cystic Lesions with a Focus on Malignant Progression

“…When EUS is coupled with needle-based confocal laser endomicroscopy (nCLE), it can provide real-time imaging of the internal epithelium of pancreatic cysts and provide an excellent diagnosis of PCLs with an accuracy of over 90% [ 110 , 111 , 112 ]. In addition, nCLE evaluation of the “thickness” and “darkness” of the cyst epithelia allows the grading of dysplasia and thus risk stratification of IPMNs [ 113 ], which can be further augmented by machine learning/artificial intelligence (AI) [ 114 , 115 , 116 ]. In the analysis of high-yield (edited) EUS-nCLE videos using a preliminary AI model, a higher accuracy rate of 82% was achieved for detecting advanced neoplasia compared to the AGA guidelines (68.6%) and Fukuoka criteria (74.3%) [ 116 ].…”

Basic Principles and Role of Endoscopic Ultrasound in Diagnosis and Differentiation of Pancreatic Cancer from Other Pancreatic Lesions: A Comprehensive Review of Endoscopic Ultrasound for Pancreatic Cancer

“…AI-enabled automatic visual inspection has proven to be helpful in rapid onsite tissue evaluation by indicating specific areas that are highly likely to indicate tumor cells in patients with pancreatic ductal adenocarcinoma with a sensitivity, specificity, and accuracy of about 80% [118,119]. Jiang et al showed that the accuracy of AI was 99.6% in differentiating low-versus high-grade neoplasia, and Nuon et al and Machicado demonstrated accuracies of 83% and 82% for AI models in differentiating mucinous cystic neoplasm versus serous cystadenocarcinoma and low versus high-grade dysplasia in intrapapillary mucinous neoplasm, respectively [120][121][122]. However, these findings should be interpreted in light of the fact that these studies were limited by a small sample size, usually from a single center.…”

Pancreatic Cancer Health Disparity: Pharmacologic Anthropology

“…Synergy between researchers, data scientists, and healthcare providers can harness the power of artificial intelligence (AI) and data analytics to improve PCa outcomes. AI algorithms can assist in early detection, risk prediction, treatment planning, and response monitoring [258,259]. However, challenges include the availability of diverse and representative datasets, addressing biases in AI algorithms, and integrating AI tools into clinical practice in an equitable and responsible manner.…”