Some types of localised breast cancer, including stage II and III, HER2 positive (Human Epidermal Growth Factor Receptor) or triple negative cancers are treated with primary or neoadjuvant chemotherapy prior to surgical treatment (mastectomy or conservative treatment). Follow-up with neoadjuvant or primary chemotherapy is usually done by breast Magnetic Resonance Imaging (MRI).Tumour response to treatment is assessed on morphological size criteria. Angiomammography (also called CESM : Contrast Enhanced Spectral Mammography) is an innovative and validated imaging technique consisting of dual energy mammography with injection of iodinated contrast medium; two images are generated, one comparable to a standard mammography and a second image highlighting the structures enhanced by the contrast medium. The characteristic neo angiogenesis of the tumour process thus allows good visualisation of the tumour compared to the underlying mammary gland. A classical morphological analysis is therefore possible thanks to standard mammographic acquisition coupled with a quantitative functional analysis linked to the study of enhancement. During angiomammography, several images are acquired in succession, starting with the cranio caudal view and ending with the profile view. On the same examination, these acquisitions are carried out at different injection times, making it possible to study the type of enhancement of a given area (Progressive/ Plateau/ Wash out). Angiomammography has several advantages over breast MRI in the follow-up of chemotherapy: shorter examination time, shorter appointment time, better tolerated by the patients and without injection of Gadolinium chelates, which have recently been shown to cause definitive brain deposits during repeated injections. Currently, angiomammography is validated in the follow-up of breast cancers treated with neoadjuvant or primary chemotherapy, with an analysis of tumour response currently only morphological, as in breast MRI. Here, the hypothesis is that the study of tumour enhancement by angiomammography may constitute a new predictive element of histological response: indeed, the performance of an initial angiomammography before treatment (baseline) and then a second angiomammography performed early in relation to the start of chemotherapy, could allow early prediction of which patients will have a complete histological response at the time of the closing surgery. This prospective study is to compare the variation in tumour enhancement, assessed on each of the two examinations using dedicated X-ray consoles, to the final histological result after closure surgery: it is hoped that the relative variation in tumour enhancement would be greater in responding patients.