Model-data-driven adversarial active learning for brain tumor segmentation

Siteng Ma, Prateek Mathur, Zheng Ju, Aonghus Lawlor, Ruihai Dong

Abstract: Highlights•Active learning for brain tumor segmentation task.•Tackles class imbalance, high domain differences, and data scarcity challenges.•Combines model impact and data stability via adversarial attacks.•Fewer samples to achieve competitive results with full supervision.•Boosts development of resource-efficient medical segmentation techniques.

External IDs:dblp:journals/cbm/MaMJLD24