Use of positron emission tomography (PET) in determination of functional tumour volume can provide usable information for radiotherapy to define the irradiated volume.

To determine the best tumour volume measure method, the investigators have chosen as model the breast cancer which allows us to study a stationary or moving organ by the patient position and belonging to a primary surgery.

The used methodology is based on lesion volume measure in preoperative time, obtained with PET, and on the measure of specimen volume by the anatomic laboratory after surgery.

This study's main objective is to compare this two measure of tumour size and secondarily to compare TEP with or without respiratory gating.

The PET-scan is achieved with FDG, under his French permission marketing, and acquire in 3 times:

A whole body acquisition in supine position, follow by a centered tumour acquisition with respiratory gating, then an acquisition in prone position to immobilise the lesion.

This study is monocentric and descriptive. It provides to include 30 patients in 1 year.

The use of positron emission tomography (PET) in determination of functional tumour volume may pose two major problems especially for the exact delineation of tumour contours:

• Respiratory movements, when the tumour is thoracic, may induce an overestimation
• The PET's low spatial resolution, linked to the emission of photons

The usual method to define tumour contours result of three types of delineation usually used: the visual contouring, the segmentation based on an activity threshold fixed, and the segmentation with adaptive thresholding.

Inclusion Criteria:

• Invasive non-lobular breast cancer, histologically proven by biopsy
• Tumour T2 - T3 regardless of nodal status, stemming from a primary surgery, in whom FDG-PET is performed for staging
• Age greater than or equal to 18 years
• Good condition: WHO lower 2.
• Women receiving effective contraception throughout the duration of treatment and 3 months after cessation of treatment,
• Patient has signed informed consent

Exclusion Criteria:

• Carcinoma other than breast ductal carcinoma.
• Patients for whom no target tumor is assessable.
• Patients for whom surgery first of their breast cancer is not indicated (tumor spread, metastasis, general health, co-morbidities).
• Pregnant women, or likely to be during breastfeeding.
• Major protected (under supervision and under guardianship)
• Failure to submit to medical supervision of the study for reasons of geography, social or psychological
• Incomplete tumor resection

patient with invasive non-lobular breast cancer, histologically proved