You are viewing the site in preview mode

Skip to main content

Table 11 Comparison with recent findings

From: DeepLeish: a deep learning based support system for the detection of Leishmaniasis parasite from Giemsa-stained microscope images

Author

Method

Types datasets

Results Obtained

P. A. Nogueira et al. [18]

Otsu thresholding; SVM classifier

Fluorescence microscopic image

85.3% classification accuracy

M. Farahi et al. [19]

Preprocessing (contrast stretching, masking); Segmentation (Modified Chan-vese (CV) Level Set Method)

Geimsa stained Microscopy images

Segmentation error of 10.9% using global CV and 9.76% using local CV

J. C. Neves et al. [20]

Blob detection based classification

Fluorescence microscopic image

Average F1 score of 84.48%

F. Ouertani et al. [21]

Watershed segmentation technique combined with region merging

Fluorescence microscopic image

NA

M. Górriz et al. [16]

U-net model classification

Florescence-stained microscopic image

NA

E. Yazdanparast et al. [24]

INsPECT (infection level measurement); Preprocessing (morphological filtering)

Fluorescent DNA staining

NA

M. Zare et al. [22]

Machine learning based Leishman parasite detection

Geimsa-stained microscope images

Precision 50% recall 60%

Proposed Model

Object detection with YOLO_V5

Geimsa-stained microscope images

MAP 73%, Precision 68%, Recall 69%