Mechanical thrombectomy has become the standard treatment for acute ischemic stroke caused by large vessel occlusion (LVO). Etiological diagnosis is increasingly a critical determinant for endovascular strategy selection, postoperative antithrombotic regimens, and long-term secondary prevention. The conventional Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification often faces limitations in emergency settings, including overlapping mechanisms, incomplete diagnostic workflows, and inconsistent levels of evidence quality, making it difficult to meet the rapid and reproducible clinical requirements. The ischemic stroke phenotyping system 2025 (ISPS25) emphasizes hierarchical evaluation of etiological evidence using a standardized diagnostic checklist, incorporating underrecognized pathogenic mechanisms (such as occult atrial fibrillation, cancer-related hypercoagulability, and arterial dissection) into a unified framework, thereby providing a more real-world paradigm for etiological assessment of acute LVO and future trial stratification. This review focuses on the major etiological subtypes of LVO and systematically summarizes key evidence for etiological differentiation from multiple dimensions, including clinical-imaging indicators, intraoperative digital subtraction angiography characteristics, artificial intelligence and machine learning-based multimodal inference, and blood- and thrombus-related biomarkers. It further discusses how these factors influence thrombectomy technique selection, rescue balloon angioplasty or stenting strategies, and postoperative antithrombotic protocols. Future research should integrate quantitative features from computed tomography (CT)/CT angiography/CT perfusion, clinical variables, laboratory indicators, and intraoperative parameters under standardized frameworks such as ISPS25, so as to construct an intelligent etiological diagnostic model with rapid inference capability, interpretable outputs and multicenter generalizability, and validate its clinical value in treatment-strategy matching, reducing reocclusion and complication risks, and optimizing secondary prevention strategies via prospective multicenter studies.