Accurate leukocyte detection based on deformable-DETR and multi-level feature fusion for aiding diagnosis of blood diseases

Yifei Chen, Chenyan Zhang, Ben Chen, Yiyu Huang, Yifei Sun, Changmiao Wang, Xianjun Fu, Yuxing Dai, Feiwei Qin, Yong Peng, Yu Gao

Abstract: Highlights•Introducing MFDS-DETR: A novel method for leukocyte detection, combining multi-scale fusion and deformable self-attention. Components: Backbone Network, HS-FPN, Encoder, and Decoder.•HS-FPN: An innovative fusion module tailored for leukocytes, addressing scale gaps in the detection dataset, enhancing feature expression.•MFDS-DETR outperforms rivals in leukocyte detection on LISC, BCCD, and WBCCD datasets, serving as a baseline.•Our team, collaborating with doctors, labeled LISC and created WBCCD dataset, crucial for leukocyte detection development. Download link on GitHub for researchers.