A Deep Learning Framework for Leukemia Cancer Detection in Microscopic Blood Samples Using Squeeze and Excitation Learning

Table 2

Results of the proposed model on both ALL1_IDB1 and ALL_IDB2.

Exp#	Class-wise performance	Run#	Dataset	Accuracy (%)	Precision (%)	Recall (%)	FScore (%)
01	ALL	01	ALL_IDB1	100	100	100	100
02	Not ALL	01	ALL_IDB1	100	100	100	100
03	ALL	01	ALL_IDB2	96	96	96	96
04	Not ALL	01	ALL_IDB2	96	97	97	97
05	ALL	02	ALL_IDB1	100	100	100	100
06	Not ALL	02	ALL_IDB1	100	100	100	100
07	ALL	02	ALL_IDB2	98	100	96	98
08	Not ALL	02	ALL_IDB2	98	96	100	98
09	ALL	03	ALL_IDB1	100	100	100	100
10	Not ALL	03	ALL_IDB1	100	100	100	100
11	ALL	03	ALL_IDB2	99.98	99 .03	99.87	99.44
12	Not ALL	03	ALL_IDB2	99.98	99.24	99.63	99.43
Results by integrating both datasets i-e ALL_IDB1 and ALL_IDB2
12	Not ALL	01	ALL_IDB1 + ALL_IDB2	97.06	97.12	97.01	97.06
13	ALL	01	ALL_IDB1 + ALL_IDB2	97.06	97.03	97.21	97.11
14	Not ALL	02	ALL_IDB1 + ALL_IDB2	99	100	97.00	99.00
15	ALL	02	ALL_IDB1 + ALL_IDB2	99.24	97.00	100	99.00
16	ALL	03	ALL_IDB1 + ALL_IDB2	99.33	99.3	99.24	99.26
17	Not ALL	03	ALL_IDB1 + ALL_IDB2	99.01	99.36	99.00	99.17