An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Parasites & Vectors

Table 2 Performance comparison of the proposed YOLO model with those reported other published works

Authors	Techniques	Datasets/Blood smears	Plasmodium parasites	mAP (%)
Yang F et al. [30]	Modified YOLOv2	Self-collected 2567 thin blood smear images	P. vivax	79.22
Koirala. et al. [33]	Modified YOLOv3 and YOLOV4 Tiny	Self-collected 3885 thick blood smear images	P. falciparum	94.07
Sukumarran et al. [32]	YOLOv4	MP-IDB and 236 images from MRC-UNIMAS Sarawak	P. falciparum P. vivax P. ovale P. malariae P. knowlesi	84.04
Abdurahman et al. [62]	Modified YOLOv3 and YOLOv4	Publicly available 1182 thick blood smear images	P. falcipraum	89.73
Present study (YOLOv4-RC3_4)	Modified YOLOv4	MP-IDB The malaria parasite image database and new dataset from MRC-UNIMAS Sarawak	P. falciparum P. vivax P. ovale P. malariae P. knowlesi	90.07

mAP Mean average precision, MP-IDB Malaria Parasite Image Database, MRC-UNIMAS Malaria Research Centre-Universiti Malaysia Sarawak, YOLO You Only Look Once object detection algorithm

ISSN: 1756-3305