Towards a precise test for malaria diagnosis in the Brazilian Amazon: comparison among field microscopy, a rapid diagnostic test, nested PCR, and a computational expert system based on artificial neural networks

Table 6 Overall performance to discriminate asymptomatic malaria cases

Diagnostic method	Correct diagnosis (%)§**	True positive (%)**	True negative (%)***	False positive (%)***	False negative (%)**
Microscopy	61.25	22.5	100	0	77.5
MalDANN Epidemiologic	56	70	28	72	30
MalDANN Epidemiologic + cytokines	80	67.5	92.5	7.5	32.5

MalDANN: Malaria Diagnosis by Artificial Neural Networks. This diagnostic software was trained on the sample of 300 individuals actively screened for asymptomatic malaria and was validated in 80 other individuals, according to the methods. The diagnostic methods presented significantly different results estimated using the Chi-square test. §correct diagnosis involves both negative and positive correct exams. **p < 0.01; ***p < 0.0001.

ISSN: 1475-2875