Journal of Medical Engineering / 2015 / Article / Tab 6

Research Article

Detection and Classification of Cancer from Microscopic Biopsy Images Using Clinically Significant and Biologically Interpretable Features

Table 6

The comparison of the proposed method with other standard methods.

Authors (year)Feature set used Methods of classificationParameters used (%)Dataset used

Huang and Lai (2010) [15]Texture features Support vector machine (SVM)Accuracy = 92.81000 × 1000, 4000 × 3000, and 275 × 275 HCC biopsy images

Di Cataldo et al. (2010) [45]Texture and morphologySupport vector machine (SVM)Accuracy = 91.77Digitized histology lung cancer IHC tissue images

He et al. (2008) [46]Shape, morphology, and textureArtificial neural network (ANN) and SVM Accuracy = 90.00Digitized histology images

Mookiah et al. (2011) [47]Texture and morphology Error backpropagation neural network (BPNN) Accuracy = 96.43, sensitivity = 92.31, and specificity = 8283 normal and 29 OSF images

Krishnan et al. (2011) [48]HOG, LBP, and LTELDAAccuracy = 82Normal-83
OSFWD-29

Krishnan et al. (2011) [48]HOG, LBP, and LTESupport vector machine (SVM) Accuracy = 88.38Histology images Normal-90
OSFWD-42
OSFD-26

Caicedo, et al. (2009) [8]Bag of features Support vector machine (SVM)Sensitivity = 92
Specificity = 88
2828 histology images

Sinha and Ramkrishan (2003) [17]Texture and statistical featuresNNAccuracy = 70.6Blood cells histology images

The proposed approach Texture, shape and morphology, HOG, wavelet color, Tamura’s feature, and LTEKNNAverage: accuracy = 92.19, sensitivity = 94.01, specificity = 81.99, BCR = 88.02, F-measure = 75.94, MCC = 71.742828 histology images