Table 6.

The comparison of our method with other works; P (precision), S (sensitivity), F1 (F1-score).

Dataset	Method	Lymph			Mon			Neut			Eos			Bas
		P (%)	S (%)	F1 (%)	P (%)	S (%)	F1 (%)	P (%)	S (%)	F1 (%)	P (%)	S (%)	F1 (%)	P (%)	S (%)	F1 (%)
LISC	Rezatofighi and Soltanian-Zadeh7	100	93.10	96.43	92	95.83	93.88	93.33	100	96.55	100	94.74	97.30	90.74	98	94.23
	Jung et al.22	100	88.89	94.12	100	66.67	80	94.44	100	97.14	100	100	100	73.91	100	85
	Baydilli and Atila23	100	100	100	80	80	80	91.66	100	96.65	100	83	90.71	88.89	88.89	88.89
	Our method	84.21	88.89	86.49	84.62	73.33	78.57	94.44	100	97.14	100	100	100	100	100	100
BCCD	Banik et al.24	99	100	99.50	100	81	89.50	74	97	83.95	96	84	89.60	–	–	–
	Liang et al.25	100	100	100	96	80	87.27	78	92	84.42	93	91	91.99	–	–	–
	Banik et al.26	100	100	100	99	99	99	93	92	92.50	93	93	93	–	–	–
	Togacar et al.19	97.99	99.80	98.89	95.83	100	97.87	93.49	92.09	92.78	94.92	90.86	92.85	–	–	–
	Our method	100	100	100	100	100	100	94.12	97.96	96	88.89	72.73	80	–	–	–

Bold values Illustrate the best-obtained value.