Deep learning-based image annotation for leukocyte segmentation and classification of blood cell morphology

BMC Medical Imaging

Table 3 Comparison of the proposed model with existing state-of-art models

Reference	Technique Used	Segmentation	Classification	Performance Parameters
Reference	Technique Used	Segmentation	Classification	Accuracy (%)	Precision (%)	Sensitivity (%)	F1-Score
Liang et al. [21]	CNN-RNN	×	✓	95.42	91.75	96.91	89.0
Pang et al. [22]	Fusion of Shallow and Deep Feature Maps	×	✓	94.33	91.25	96.00	89.0
Yu et al. [23]	Deep Neural Network	×	✓	90.51	81.5	92.39	81.0
Banik et al. [24]	Feature Map Fusion of Two Convolution Layers	×	✓	96.01	97	99.03	92.0
Proposed Model	Deep Feature Map Extraction of Segmented Leukocyte	✓	✓	97.98	97.97	97	97.0

ISSN: 1471-2342