2D/3D Mammography, Density Assessment Software

A mammogram procedure is a low-dose x-ray of the breast.

A screening mammogram is performed at regular intervals to check for breast cancer when there are no signs or symptoms of disease. Screening mammograms are usually reviewed after the individual leaves the breast imaging center. Results will usually be available within a few days, and, in the USA, always within 30 days.

A diagnostic mammogram is used to check for breast cancer when there is a sign or symptom of disease, or to evaluate an abnormality seen on a screening mammogram. Common symptoms include a lump, nipple discharge, skin or nipple retraction, or a change in the size or shape of the breast. Symptoms can be due to breast cancer, but are more often due to benign (noncancerous) conditions. Breast pain, in particular, is very rarely due to cancer unless there is also a lump (and tender lumps are not usually due to cancer).

A diagnostic mammogram is monitored by the radiologist at the time of the examination. Diagnostic mammography starts with the same images as a screening mammogram and may also include additional images to review an area of concern. Ultrasound may be performed as part of diagnostic imaging at the same visit. Results of a diagnostic mammogram are given immediately to the patient.

2D Digital Mammography: The breast is placed on the surface (detector) of the mammography system and is briefly compressed between two paddles while an x-ray is taken (Fig. Mammo-1). X-rays of each breast are taken from two different positions to make sure the maximum amount of tissue is included. Occasionally additional images may be needed to fully include all the breast tissue. The total examination takes about 10 minutes. Compression spreads the breast tissue out to reduce tissue overlap and also reduces the radiation dose. Breast compression also limits motion to produce clearer images.

3D mammography (tomosynthesis): A standard 2D digital mammogram provides a 2-dimensional picture of the breast, which is a 3-dimensional object. The normal structures of the breasts (such as milk glands, fibrous tissue, fat, and blood vessels) overlap one another when viewed as a 2D image. Overlapping tissue can hide breast cancers and sometimes makes normal tissue look like breast cancer.

3D/tomosynthesis images are taken at multiple angles. The images are then reconstructed as multiple thin “slices” which can be individually scrolled through by the radiologist to review (Figs. Mammo-2, 3), like scrolling through pages of a book. Tomosynthesis can help reduce the overlap of normal tissues. This improves cancer detection and lowers the chance that normal tissue will be mistaken for cancer.

When a 2D and 3D/tomosynthesis mammogram are performed together (in “combination mode”), the radiation dose to the breast is about twice that of a 2D mammogram alone – and the dose is greater in thicker or denser breasts. This is still well below the background level of radiation from living on Earth. Many centers have the computer software needed to create a “synthetic” 2D mammogram from the same images used to create the 3D/tomosynthesis slices. This synthetic mammogram can replace the standard 2D digital mammogram so that the radiation dose from 3D/tomosynthesis is similar to a 2D digital mammogram.

Why does breast density matter? The denser the breast tissue, the greater the risk of both developing breast cancer and of hiding cancer on a mammogram.

Radiologists have traditionally determined breast density by looking at mammogram images and comparing the amount of fibroglandular tissue (white/light) vs. fatty (gray) areas on a mammogram. Breast density is described as one of four Breast Imaging Reporting and Data System (BI-RADS^®) categories: A) fatty, B) scattered fibroglandular density, C) heterogeneously dense, or D) extremely dense. Breasts that are heterogeneously dense or extremely dense are considered “dense.” Breasts with fatty or scattered fibroglandular density are “not dense”. Density assessment can also be done by computer software (see explanation of Density Assessment Software below).

In the USA, beginning September 10 2024, all mammography facilities are required to include in the patient results letter whether the breasts are dense, or not dense. In the USA and some other countries, the report sent to the referring doctor or health professional will contain the patient’s specific density category. The specific density category can be found in the medical record and formal report sent to the doctor.

Breast density is included in many breast cancer risk calculators and influences recommendations for screening in addition to mammography such as with MRI or ultrasound. Breast density may change over time and should be assessed and reported on every mammogram.

Assessment of density can vary between radiologists, especially for breasts that are on the borderline between scattered fibroglandular (“not dense”) and heterogeneously dense  (“dense”­) [1]. These changes in density assessment may affect what screening is recommended in addition to mammography. Modern computer programs, such as artificial intelligence, can accurately and reproducibly calculate breast density  [2-5].  Computer measurements can provide more detail and more consistent results than human radiologists.  Computer software can identify specific areas where a cancer may be hidden and can also predict the likelihood of developing breast cancer. These assessments can be used together with a radiologist’s assessment to improve patient care [6, 7].

Breast density assessment software is costly and is not separately reimbursed by insurance carriers, nor are radiologists required to use it. Visual density assessment and software approaches show similar performance in identifying risk of screen-detected and interval cancer [8, 9].

The vast majority of invasive breast cancer is invasive ductal cancer (IDC), but about 10-15% of all breast cancer is invasive lobular cancer (ILC). ILC is especially hard to see on mammograms because it grows as single-file columns of cells and tends not to form a lump or have associated calcifications (white specks) [10]. 3D/tomosynthesis improves detection of ILC [11].

For every 1000 individuals screened, about 5 will be found to have cancer on mammography (see Summary of Cancer Detection Rates). Screening mammography is the only breast cancer screening technology that has ever been studied in large trials, known as randomized controlled trials (RCTs), to learn if it reduces the chance of dying from breast cancer. In multiple RCTs performed from the 1960s to the 1990s, mammography screening programs reduced deaths from breast cancer by 15-22% [12]. Mammogram technology has since  improved, and today mammograms reduce the risk of death from breast cancer by 40-60% in people who have them [13, 14]. The reduction in deaths is because mammograms can find cancers earlier than if a person did not get a mammogram.

Adding 3D/tomosynthesis to 2D digital mammography increases cancer detection by 1 to 2 cancers per thousand individuals screened. Several studies [15, 16] have shown there is a benefit to having 3D/tomosynthesis every year as it improves cancer detection. In studies done in the USA, 3D/tomosynthesis also reduces the number “call-backs” or “recalls” for additional testing of findings that turn out not be cancer (known as false positives)[17]. European-based studies have not shown a reduction in false positives [17, 18], likely due to low recall rates in Europe even with 2D mammograms.

All mammograms use x-ray technology. Dense tissue absorbs more x-rays than non-dense, (fatty) tissue and appears white on a mammogram. Breast cancers are also dense and appear white. Some breast cancers are hidden (masked) on a mammogram by overlying or surrounding dense breast tissue (Fig. Mammo-4). A cancer masked on a 2D mammogram can still be masked on 3D/tomosynthesis unless the cancer is at least partially surrounded by fatty tissue. Digital 2D mammography has been shown to miss about 40% of cancers present in individuals with extremely dense breasts and 25% of cancers present in individuals with heterogeneously dense breasts [19-24]. Compared to 2D mammography, 3D/tomosynthesis improves cancer detection and reduces recalls the most on the baseline (first) screening examination, though benefits continue every year [15, 16, 25]. In individuals with extremely dense breasts, there is less benefit from 3D/tomosynthesis and cancers are often still hidden [18, 26-28].

Last update 11/17/23