BADDI: Artificial intelligence in screening with tomosynthesis

The BADDI-project will add knowledge about the use of artificial intelligence to detect breast cancer in screening with tomosynthesis.

Last updated: 2/18/2025

Background

BADDI stands for Breast Cancer, Artificial Intelligence, Digital Breast Tomosynthesis, Digital Mammography and Interval Cancer.

Artificial intelligence (AI) have shown promising results in several areas of mammographic screening. However, much is still unknown about the advantages and disadvantages of such use, and there are few studies on AI in screening with tomosynthesis. Tomosynthesis is a 3D-like and more advanced type of mammography than standard digital mammography. The To-Be studies - the Tomosynthesis studies in Bergen - have investigated whether tomosynthesis is a suitable screening technique in BreastScreen Norway. The BADDI-project is based on data from these studies, specifically from To-Be 2.

Interval cancers are breast cancers detected between two screening examinations, and are an inevitable part of a screening program. Screening programs aim to keep the proportion of interval cancers low, and it is interesting to investigate whether AI systems can detect breast cancers that are particularly difficult to detect for radiologists.

This project will increase our knowledge about which tumors AI can detect on tomosynthesis images. The project will also provide knowledge about whether AI can detect more breast cancers on tomosynthesis images than radiologists, without increasing the proportion of false-positive screening examinations and without finding more small, slow-growing breast cancer tumors that may increase the risk of overdiagnosis and overtreatment of women.

Purpose

The purpose of the BADDI-project is to increase our knowledge about the ability AI has to detect breast cancers in screening with tomosynthesis, and to investigate whether AI systems can be as good or better at detecting breast cancers than radiologists.

Data

BADDI will use information from screening examinations already conducted in the To-Be studiescarried out in BreastScreen Norway in the period 2016-2019. We will use information from the tomosynthesis and mammography images (image data) analyzed by AI, in combination with information from the radiologists' interpretations and results from the screening examinations (screening information).

In To-Be 2, about 31 000 women were screened, and among these about 320 breast cancers were detected, including about 50 interval cancers. These women form the basis of the study population in BADDI.

BADDI is a register-based study, and the women will not be contacted about the project. It will be impossible to recognize individuals in published results.

Study plan

The study will use an AI system to analyze To-Be 2 tomosynthesis images as well as previous mammograms from up to three previous screening rounds, to assign a breast cancer risk score. The goal is to investigate the sensitivity and specificity of the AI system.

We aim to explore 1) detection rates of screen-detected breast cancer and interval cancer as a result of AI analysis of To-Be 2 tomosynthesis images using different breast cancer risk score thresholds, with and without the use of previous screening images, and 2) mammographic findings and histopathological tumor characteristics associated with the cancers detected or missed by the AI system.

Organization

BADDI is a collaboration between the Cancer Registry of Norway at NIPH, the breast center at Haukeland University Hospital and the Mohn Medical Imaging and Visualization Center at Haukeland University Hospital (MMIV). NIPH is data controller and resonsible research institution. We are responsible for obtaining the necessary approvals and for extracting information about the women in the study population, and transferring this to MMIV. NIPH will also be responsible for performing analyzes, and interpreting and publishing the results.

MMIV at Haukeland University Hospital is responsible for extracting, pseudonymizing and analyzing image data using the AI system. They will also contribute in the interpretation and publication of results.

ScreenPoint Medical B.V. has developed the AI system to be used in this study, Transpara ™. The system is CE marked. ScreenPoint Medical is responsible for access, installation, calibration and training the staff in how to use the system.

Status

As of February 2025, image data from women screened with tomosynthesis in To-Be 2, as well as with digital mammography or tomosynthesis at up to three previous screening appointments in BreastScreen Norway, has been extracted. An updated version of the AI system that can analyze tomosynthesis images was ready for use in January 2025, and MMIV has now started analyzing the tomosynthesis images.

About the project

Project leader: Solveig Hofvind

Postdoc: Nataliia Moshina

Project group: Åsne Holen, Marthe Larsen

Funding

The project is financed by the Cancer Society through Open Call 2020, the Cancer Registry of Norway and Haukeland University Hospital.

Duration

1.1.2021-31.12.2025

Data protection

The project is approved by the Regional Comittees for Medical and Health Research Ethics, project number 2015/424 and 13294).

You can find general information about data protection in the Cancer Registry of Norway here.

Collaborators

External collaborator is Mohn Medical Imaging and Visualization Centre at Haukeland University Hospital, by professor Ingfrid Haldorsen.