Rivers are among the most dynamic components of the surface water cycle and play a vital role in the global hydrological system. Under the combined pressures of climate change and human activities, river meander migration has intensified, leading to localized channel cutoff events with significant ecological and management implications. Although numerous studies have documented changes in river migration, they have predominantly addressed local-scale channel dynamics, and automated methods for detecting large-scale channel cutoff events are still lacking. In this study, we propose a scalable method that integrates multi-temporal river-extent mapping with morphological analysis to detect channel cutoffs across broad spatial domains. Our technique identifies closed loops formed by intersecting river centerlines from different periods, thereby reducing false positives from bifurcations and enhancing computational efficiency. We applied the method to five major rivers, the Amazon, Tarim, Yellow, Red, and Ganges, spanning diverse climatic, hydrological, and geomorphic settings. Validation against manual inventories yielded a median F1-score of 0.86 (with a peak of 0.90 for the Yellow River), demonstrating high accuracy and robustness. Cutoff frequency and morphology varied markedly: the hyper-arid Tarim River exhibited the highest frequency (9.2 events per 100 km), with a limited spatial extent, while the Amazon showed a lower frequency (0.6 events per 100 km) but larger events (intersection lengths >2 km; loop areas >0.8 km(2)). This study presents a generalizable and computationally efficient framework for monitoring river cutoff dynamics, providing a robust foundation for advancing our understanding of river evolution and informing sustainable river management worldwide.