Table 5. Characteristics of deep learning systems for CVD risk predictions (III).

SN	Studies	#PE P	Sen	Spec	Acc	PPV	NPV	FPR	Pre	F1 Score	P value	HL	D Coff.	JI
1	Azzorpardi et al. (78)	3	√	√	×	×	×	×	×	×	√	×	√	×
2	Biswas et al. (79)	3	×	×	√	×	×	×	×	×	√	×	√	×
3	Biswas et al. (80)	4	√	√	95.4	×	×	×	×	×	×	×	√	×
4	Biswas et al. (81)	3	O	O	99	×	×	×	×	×	√	×	√	×
5	Jain et al. (98)	7	√	√	√	×	×	×	√	×	√	×	√	√
6	Jain et al. (99)	7	√	√	√	×	×	×	√	×	√	×	√	√
7	Jamthikar et al. (13)	3	√	√	√	×	×	×	×	×	×	×	×	×
8	Lakadir et al. (82)	3	83	90	78.5	×	×	×	×	×	×	×	×	×
9	Meshram et al. (83)	3	83	√	O	×	×	×	×	×	×	×	√	×
10	Wu et al. (84)	3	√	√	89	×	×	×	×	×	×	×	√	×
11	Zhou et al. (85)	5	√	√	√	×	×	×	√	×	√	×	×	×
12	Zhou et al. (86)	6	√	√	√	×	×	×	√	×	√	×	√	×
13	Ganitidis et al. (87)	3	75	70	75	×	×	×	×	×	×	×	×	×
14	Mohannadi et al. (100)	3	√	√	98	×	×	×	×	×	×	×	√	×
15	Latha et al. (101)	5	√	√	100	×	×	×	×	×	×	×	√	√
16	Otgonbaatar et al. (102)	3	√	√	√	×	×	×	×	×	×	×	×	×
17	Jain et al. (103)	6	√	√	√	×	×	×	√	×	×	×	√	√
18	Ziegler et al. (88)	7	√	√	√	×	×	×	√	×	√	×	√	√
19	Bortsova et al. (89)	3	83.8	×	×	88	×	×	×	×	√	×	×	×
20	Zhu et al. (104)	3	×	×	×	×	×	×	×	×	×	√	√	√
21	Park et al. (105)	3	√	√	√	×	×	×	×	×	×	×	×	×
22	Jain et al. (90)	4	×	×	√	×	×	×	×	×	×	√	√	√
23	Savaş et al. (91)	6	×	×	√	√	√	√	√	√	×	×	×	×
24	Washim et al. (92)	1	×	×	√	×	×	×	×	×	×	×	×	×
25	Sudha et al. (93)	2	√	×	×	×	×	×	×	×	×	√	×	×
26	Groves et al. (106)	4	×	×	√	×	√	√	×	×	×	√	×	×
27	Saba et al. (94)	3	×	×	√	×	×	√	×	×	×	×	×	√
28	Tsakanikas et al. (107)	6	×	×	√	√	×	×	×	×	×	×	×	√
29	Koktzoglou et al. (95)	6	√	×	√	×	×	√	√	×	√	×	×	√
30	Flores et al. (108)	6	√	√	√	√	×	×	×	√	√	×	×	×
31	Luo et al. (96)	1	×	×	√	×	×	×	×	×	×	×	×	×
32	Xiao et al. (97)	3	√	×	√	√	×	×	×	×	×	×	×	×

CVD, cardiovascular disease; SN, serial number; # PE P, number of PE parameters; Sen, sensitivity; Spec, specificity; Acc, accuracy; PPV, positive predictive value; NPV, negative predictive value; FPR, false positive rate; Pre, precision; HL, Hamming loss; D Coff., DICE coefficient; JI, Jaccard-index.