Evaluation of methods for detection of fluorescence labeled subcellular objects in microscope images

BMC Bioinformatics

Table 2 Results for yeast image frames

Algorithm	precision	recall	F-score
BPF	0.9570	0.9351	0.9459
FPD	0.5964	0.8969	0.7165
HD	0.8682	0.7290	0.7925
KDE	0.9116	0.8664	0.8885
LC	0.9396	0.9504	0.9450
LEF	0.8712	0.8779	0.8745
MGI	0.6175	0.8626	0.7198
MW	0.7645	0.8550	0.8072
SE	0.9318	0.9389	0.9354
SPL	0.8167	0.9351	0.8719
THE	0.0062	0.9733	0.0123

Summary of numerical results for the frames from image stacks of yeast cells. The reported results are the precision, recall, and F-score values calculated for 28 frames, seven out of which had no objects according to manual analysis. The results are the maximum F-scores obtained in parameter tuning by grid search, i.e., the F-score shown with red dots in Fig. 3.

ISSN: 1471-2105