Detection of Peripheral Malarial Parasites in Blood Smears Using Deep Learning Models

Table 4

Proposed CNN with state-of-the-art models, 96%.

Authors work	Methodology	Accuracy in %	Sensitivity	Specificity	F1-score
Rajaraman et al. [2]	Pretrained-CNN	0.98	0.981	0.992	0.987
Vijayalakshmi et al. [5]	CNN with 2-level	0.977	0.971	0.972	0.959
Liang et al. [4]	CNN (16 layers)	0.973	0.969	0.977	—
Das et al. [9]	Bayesian learning	0.84	0.981	0.689	—
Ross et al. [8]	Image enhancement	0.73	0.85
Proposed model	Customized CNN	0.98	0.985	0.988	0.987