Accurate and fast mitotic detection using an anchor-free method based on full-scale connection with recurrent deep layer aggregation in 4D microscopy images

Table 2 Results of a quantitative comparison among the proposed network, non-deep learning methods, and deep learning methods

Method	Precision	Recall	F1 score
EDCRF [35]	0.6829	0.6210	0.652
SVM [36]	0.3782	0.9035	0.6409
2D Faster R-CNN [17]	0.0870	0.9310	0.509
3D Faster R-CNN	0.0592	0.4143	0.2367
SSD [16]	0.0411	0.7221	0.3816
CornerNet(Hourglass)	0.0715	0.8933	0.4823
CenterNet (DLA)	0.1002	0.9018	0.5010
HCNN [6]	0.7003	0.6910	0.6957
TS-BLSTM [5]	0.7883	0.7751	0.7817
2.5D Faster R-CNN [14]	0.3591	0.7532	0.5562
CasDetNet_CNN [14]	0.7228	0.70358	0.7132
CasDetNet_CLSTM [14]	0.8195	0.7974	0.8085
CasDetNet_3DAnchor [15]	0.8356	0.8442	0.8399
CenterNet(RDLA++)	0.8339	0.8752	0.8546

ISSN: 1471-2105