Table 1 Significant features on BCDR database
From: A machine learning approach on multiscale texture analysis for breast microcalcification diagnosis
Normal/Abnormal | Benign/Malignant | ||||||
|---|---|---|---|---|---|---|---|
Embedded M. | freq (%) | Filter M. | freq (%) | Embedded M. | freq (%) | Filter M. | freq (%) |
(k ≤2) | (k ≤6) | (k ≤10) | (k ≤26) | ||||
# Interest Points | 100 | # Interest Points | 100 | Variance _LL2 | 100 | Variance _LL1 | 100 |
# Interest Corners | 100 | Kurtosis _HL2 | 99.80 | # Interest Corners | 100 | Skewness _LL1 | 100 |
# Interest Corners | 99.10 | Variance _LL1 | 99.90 | Entropy _LL1 | 100 | ||
Kurtosis _HL1 | 97.80 | RelSmoothness _LL2 | 99.90 | RelSmoothness _LL1 | 100 | ||
Kurtosis _LH1 | 76.40 | RelSmoothness _LL1 | 99.60 | Entropy _HL1 | 100 | ||
Kurtosis _LH2 | 61.90 | # Interest Points | 91.30 | Entropy _HH1 | 100 | ||
Variance _LH2 | 24.80 | Variance _HH1 | 77.70 | Kurtosis _HH1 | 100 | ||
RelSmoothness _LH2 | 21.90 | RelSmoothness _HH1 | 77.40 | Variance _LL2 | 100 | ||
Entropy _HH1 | 58.90 | Skewness _LL2 | 100 | ||||
Entropy _HL1 | 44.80 | Entropy _LL2 | 100 | ||||
Mean _HH1 | 41.20 | RelSmoothness _LL2 | 100 | ||||
Kurtosis _LH2 | 100 | ||||||
Kurtosis _HL2 | 100 | ||||||
Kurtosis _HH2 | 100 | ||||||
# Interest Points | 100 | ||||||
# Interest Corners | 100 | ||||||
Entropy _LH1 | 99.20 | ||||||
Entropy _LH2 | 98.60 | ||||||
Entropy _HH2 | 97.80 | ||||||
Kurtosis _HL1 | 97.10 | ||||||
RelSmoothness _HH1 | 96.10 | ||||||
Variance _HH1 | 88.80 | ||||||
Skewness _HL2 | 76.30 | ||||||
Mean _LL1 | 59.00 | ||||||