Interval symbolic feature extraction for thermography breast cancer detection

• Malignant, benignant and cyst classes are classified using a symbolic feature extraction.
• Features based on interval dissimilarities are obtained from breast thermograms.
• The proposed feature extraction surpassed statistical and texture feature extractions.
• The results of performance were 85.7% of sensitivity index for the malignant class and accuracy of 84%.

Breast cancer is one of the leading causes of death in women. Recent studies involving the use of thermal imaging as a screening technique have generated a growing interest especially in cases where the mammography is limited, as in young patients who have dense breast tissue. The aim of this work is to evaluate the feasibility of using interval data in the symbolic data analysis (SDA) framework to model breast abnormalities (malignant, benign and cyst) in order to detect breast cancer. SDA allows a more realistic description of the input units by taking into consideration their internal variation. In this direction, a three-stage feature extraction approach is proposed. In the first stage four intervals variables are obtained by the minimum and maximum temperature values from the morphological and thermal matrices. In the second one, operators based on dissimilarities for intervals are considered and then continuous features are obtained. In the last one, these continuous features are transformed by the Fisher’s criterion, giving the input data to the classification process. This three-stage approach is applied to a Brazilian’s thermography breast database and it is compared with a statistical feature extraction and a texture feature extraction approach widely used in thermal imaging studies. Different classifiers are considered to detect breast cancer, achieving 16% of misclassification rate, 85.7% of sensitivity and 86.5% of specificity to the malignant class.