Automatic Detection of Cervical Cancer Cells by a Two-Level Cascade Classification System

Table 1

Example morphologic and texture features of the nuclei of different type of cells.

	Step 1 classification				Step 2 classification
Cell type	Epithelial cells	Lymphoid cells	Neutrophils cells	Junk cells	Normal epithelial cells	Abnormal epithelial cells
Cell image
Area ()	1071	403	342	972	538	2455
Circularity	0.843519	0.916002	0.765056	0.300962	0.753142	0.997429
Distance	17.397	10.2599	9.30611	15.8119	12.0604	26.8547
Sigma	1.16761	0.546327	1.61327	5.60155	1.2382	0.346338
Roundness	0.932884	0.946751	0.826644	0.645737	0.897334	0.987103
Sides	5.05552	5.63962	3.22911	2.30386	4.1358	11.0196
Mean	158.233	79.5434	100.567	128.506	117.796	123.245
Deviation	6.99297	41.8068	39.1483	28.4216	29.373	37.8975
Energy	0.0744681	0.00502435	0.00584795	0.0033976	0.00580941	0.0416018
Correlation	0.872788	0.891784	0.901872	0.852879	0.937312	0.93632
Homogeneity	0.716823	0.301341	0.281612	0.258799	0.368374	0.574155
Contrast	0.868347	25.5558	19.7018	15.9352	7.09665	12.4264
Convexity	0.97541	0.975787	0.914439	0.675939	0.967626	0.983574
	105447	13084.7	12303.9	72230.5	18714	480713
	79455.2	12887.3	7861.95	131805	28798.7	478728
	106538	13115.8	12645.8	131821	29095.8	480881
	78364.5	12856.2	7520.04	72215.2	18416.8	478559
Entropy	4.58423	6.75867	6.62043	6.56861	6.61984	5.98287
Anisotropy
Compactness	1.1056	1.04565	1.30927	5.38011	1.11106	1.10762
ContLength (μm)	121.983	72.7696	75.0122	256.35	86.669	184.853
Diameter (μm)	39.3954	23.2594	23.7697	57.8705	29.4279	55.9464
Radius (μm)	16.5	10.5	7	9	11.5	27
Rectangularity	0.809384	0.826425	0.760234	0.579734	0.77907	0.807594
Anisometry	1.16598	1.01004	1.29677	1.35107	1.25692	1.00242
Bulkiness	1.00102	1.00474	1.04771	1.29773	1.00501	1.00022
Structure-Factor	0.16717	0.0148302	0.358639	0.753318	0.263212	0.002.63965