The inclusion criteria for this study were women older than 18 years with a diagnosis of TN and/or HER2+ invasive ductal breast cancer, T1-2 N0 M0, who underwent to NAT in a single institution. Patients with multifocal, multicentric, and contralateral breast cancers, diagnosis of associated ductal carcinoma in situ (DCIS), and/or presence of chronic or psychiatric disorders were excluded from this study. Imaging assessments, including digital mammography, were conducted before and after the neoadjuvant therapy. Specifically, mammograms were performed using digital mammography in bilateral cranio-caudal and medio-lateral views, as well as medio-lateral-oblique views in tomosynthesis. Four radiologists assessed the mammograms: one highly experienced reader (over 20 years in breast imaging), one moderately experienced reader (7 years in breast imaging), one radiology resident with low experience (1 year in breast imaging), and one medical doctor with very low experience (a sixth-year medical student not certified as a radiologist). The radiologists evaluated the mammograms for changes in breast tissue density using the ACR BI-RADS breast density classification. The study employed a commercial AI tool that use machine learning algorithms to perform automated measurements, providing an independent assessment that can be compared with the in-house developed AI tool and radiologists' evaluations.