Breastcancer detection Interpretability with LIME

Considering the disease of breast cancer that affects 1 in 8 women, the application of Deep Learning technique in this context becomes increasingly present believing that it is an effective way to apply faster diagnoses and thus contribute to the solution of this problem . The study model uses histopathological images to detect whether it is benign or malignant, considering the BreakHis database. Considering this problem and the importance of the application of interpretability techniques in the health field, we apply the LIME technique in this study and thus observe how the model behaves for certain examples and thus generate explanations that in the future can be validated by a specialist. For more details on building the model