Skip to main content
. 2022 Apr 8;12:5979. doi: 10.1038/s41598-022-09954-8

Table 1.

The reported metrics of the selected studies.

STUDY SET REPORTED EVALUATION TOTAL POS NEG TP TN FP FN ACC PREC REC F1 SPEC MCC NPV TS
1 1 In-text Per-finding 338 337 1 1.00
2 1 T2.R1 Per-frame 210 143 67 0.71 0.89 0.68 0.75
1 T2.R2 Per-frame 210 143 67 0.77 0.81 0.88 0.83
1 T2.R3 Per-frame 210 143 67 0.87 0.91 0.84 0.87
2 T3.R1 Per-frame 48 13 35 0.65 0.85 0.73
2 T3.R2 Per-frame 48 35 13 0.94 0.83 0.88
2 T3.R3 Per-frame 0.83 0.86 0.83 0.84
3 1 T1 Per-frame 106 65 33 7 1 0.94 0.90 0.98 0.83 0.97
4 1 T2.R1 Per-frame 6000 6000 0 5663 0 251 337 0.94
1 T2.R2 Per-frame 1414 1414 0 1296 0 41 118 0.92
1 T2.R3 Per-frame 21,572 0 21,572 0 20,691 1004 0 NA 0.96
2 T2.R4 Per-frame 570 0 42 76 0.88
3 In-text Per-frame 60,914 0.92
4 In-text Per-frame 1,072,483 0 1,072,483 0 0 0.95
5 1 T1 Per-frame 3723 4735 262 930 0.88 0.80 0.95

The STUDY column represents each of the five studies selected for metric recalculation. The SET column is the different metrics calculated for the same set of data. The REPORTED column is how the metrics were reported in the respective study. To refer to the tables in each respective paper, we use T to refer to the table number and R for the row number. The EVALUATION column is the method used to generate the metrics. The TOTAL column is the total number of samples used in the metrics calculations. The POS and NEG columns represent the total number of positive and negative samples, respectively. The remaining columns correspond the aforementioned metric acronyms described in the main text.