One of the highly dangerous along with the 2nd major diseases that cause female death is Breast Cancer
(BC). The growth of malignant, cancerous lumps as of the breast cells is the beginning stage of BC. It can be
diagnosed earlier by self-tests along with regular clinical checks; in that way enhance the survival probability
considerably. For researchers along with scientists, BC classification is a highly challenging task. In cancer data
classification, Neural Networks have turned into a well-liked tool in recent days. Illustrating the present
understanding of Convolutional Neural Networks (CNNs) in BC Detection (BCD) in an organized along with
structured format is the intention of this survey. In this, the CNN- centric Transfer Learning (TL) is investigated to
portray breast masses for various diagnostic, predictive, or prognostic tasks or in a variety of imaging modalities like
Ultrasound (US), Magnetic Resonance Imaging (MRI), Digital Mammography (DM), together with Digital Breast
Tomosynthesis (DBT). The recognition of advantages in conjunction with the drawbacks of BCD utilizing CNN is
the motive of this work. Lastly, the potential chances for future work are spotlighted in this paper. In this, a road
map for developing the CNN-centric solutions to enhance BC further is provided by the materials illustrated along
with discussed here.