Statistical Analysis on Breast Thermograms using Logistic Regression for Image Classification

Breast cancer is one of the most often diagnosed cancers in women. The only approach to lower the mortality rate is to diagnose it early and accurately. Breast thermography is a non-invasive, painless imaging modality that may be utilised in conjunction with an existing mammography imaging equipment to provide a reliable screening tool. Because mammography only detects palpable tumours, thermography combined with mammography can provide an accurate prediction. On the basis of breast thermograms, this study offers statistical analysis utilising regression. The images are first subjected to region-of-interest removal and then feature extraction. The classification accuracy is predicted using regression analysis on the obtained features and the R
²
value. The findings of logistic regression reveal that R
²
= 40.2 in variance of features in the classification outcome predicts 89 %.