Neuroendocrine neoplasms (NENs) are a diverse group of tumors characterized by the expression of neuroendocrine markers. While NENs can theoretically arise from neuroendocrine cells throughout the body, gastroenteropancreatic NENs (GEP-NENs) are without question the most common subset.1, 2, 3 Historically considered a rare disease, improved diagnostic methods and screening practices have led to a significant increase in the reported incidence of NENs, revealing a broader prevalence of these tumors than previously understood.4 NENs are notable for their prominent expression of different neuroendocrine markers. Somatostatin receptors (SSTRs) are among the most clinically relevant markers, which play a significant role in their clinical management. The expression of SSTRs is a characteristic feature of NENs that enables both diagnostic and therapeutic interventions. SSTR imaging, particularly Positron Emission Tomography/Computed Tomography (PET/CT) with radiotracer-linked peptides specific for SSTR2 (e.g. 68Ga-DOTATATE, 68Ga-DOTATOC), has emerged as a powerful tool for the diagnosis, staging, monitoring, and prognostication of NENs.5,6 Other functional imaging modalities, such as fluorodeoxyglucose (FDG) PET, can provide complementary diagnostic information to SSTR-based imaging as the increased metabolic activity detected by FDG uptake often correlates with more aggressive tumor behavior.7, 8, 9

Despite fast advancements in diagnostic technology, NENs remain challenging to diagnose and characterize due to their heterogeneous nature, which complicates classification and prognosis. The variable clinical presentation of NENs compounded by their diverse biological behavior—ranging from well-differentiated, indolent neuroendocrine tumors (NETs) to aggressive neuroendocrine carcinomas (NECs)—poses significant challenges in disease classification and prognosis. Differentiating NETs from NECs is crucial for guiding treatment strategies. The most recent WHO classification highlighted these as distinct disease processes requiring different diagnostic and therapeutic strategies.10 Distinguishing NETs from NECs remains a significant clinical challenge, which underscores the dual diagnostic hurdle in NEN diagnosis: confirming the presence of the tumor and accurately assessing its biological behavior to guide optimal clinical management. The heterogeneity, complex biological behavior, and diverse imaging patterns of NENs, coupled with their reliance on multimodal (i.e., anatomical and functional imaging) diagnostic approaches for characterization, creates an ideal scenario for the integration of artificial intelligence (AI) applications to enhance precision in detection, classification, and treatment planning. AI-driven tools have the potential to address key clinical challenges posed by the complexity and heterogeneity of NETs, including enhancing diagnostic accuracy, refining prognostic models, personalizing treatment planning, and optimizing therapeutic approaches such as peptide receptor radionuclide therapy (PRRT). This article reviews the current status and future directions for the application of AI in the diagnosis, prognosis, and treatment of NENs. AI has proven capabilities in enhancing the clinical management of NENs from multiple aspects, including analysis of serum biomarkers, pathological specimens and genomic signatures. 11 This review focuses on the impact of AI on diagnostic imaging of NENs, including both anatomical and functional imaging. The review also highlights emerging AI methodologies, their potential integration into clinical workflows, and the opportunities and challenges associated with leveraging AI to improve outcomes in NET management.