A Dataset for Breast Cancer Histopathological Image Classification
Citations
1,328 citations
720 citations
Cites background or methods or result from "A Dataset for Breast Cancer Histopa..."
...[11] introduced a dataset composed of 7,909 breast histopathological images acquired on 82 patients....
[...]
...A set of comprehensive experiments on the BreaKHis dataset proposed in [11] shows that the CNN achieves better results than the best results obtained by the other machine learning models trained with textural...
[...]
...The performance at image level is not reported in [11]....
[...]
...Based on the results presented in [11], it is undeniable...
[...]
...However, on the histopathological images assessed, LeNet classification performance were considerably inferior to our previous results reported in [11], achieving about 72% of accuracy....
[...]
471 citations
425 citations
372 citations
Cites background from "A Dataset for Breast Cancer Histopa..."
...[30, 55, 66, 186] [123, 187, 206] [36, 181, 205] Erie County [66], EEG [9], BreaKHis [210], SVEB, SVDB [186] Biomedical...
[...]
References
[...]
79,257 citations
"A Dataset for Breast Cancer Histopa..." refers background in this paper
...The principle behind ensemble methods is that a group of weak learners (in this case the decision trees) can come together to form a strong learner [32]....
[...]
47,974 citations
"A Dataset for Breast Cancer Histopa..." refers methods in this paper
...All the experiments were carried out using scikit-learn, an opensource machine learning library in Python [33]....
[...]
37,861 citations
"A Dataset for Breast Cancer Histopa..." refers background in this paper
...Differently from other linear discriminant functions, it provides the optimal hyperplane that separates two classes [31]....
[...]
28,898 citations
20,442 citations
"A Dataset for Breast Cancer Histopa..." refers background or methods in this paper
...GLCM are widely used to characterize texture images....
[...]
...Note, however, that the 200× magnification factor also shows high potential, with the best results over GLCM and PFTAS, higher than those obtained with the 40× level....
[...]
...On the GLCM, 13 Haralick parameters are computed [20]: angular second moment, contrast, correlation, sum of squares, variance, inverse difference moment, sum average, sum variance, sum entropy, entropy, difference variance, difference entropy, information measures of correlation 1, and information measures of correlation 2....
[...]
...These include the textural descriptors most commonly found in the literature, such as local binary patterns (LBP) [17], completed LBP (CLBP) [18], local phase quantization (LPQ) [19], gray-level co-occurrence matrix (GLCM) [20], threshold adjacency statistics (TAS) [21], and one keypoint descriptor, named ORB [22]....
[...]
...In our experiments, four adjacency directions 0◦, 45◦, 90◦, 135◦, and eight gray levels are used to compute the GLCM....
[...]