A New Era in Lesion Detection

Zebra is focused on helping mammographers read and diagnose mammograms faster and with increased accuracy and assurance. Zebra’s mammography solution will aid the mammographer with a more accurate reading of suspicion lesions while reducing the number of false positives that hamper the current solutions on the market today.


Breast cancer is the second most common cancer in the world, and by far the most frequent cancer among women. As early detection is key for saving lives and decrease the high mortality rate, many health systems and governmental programs are actively screening women in the target population. Through early detection in asymptomatic women, screening programs aim to reduce breast cancer mortality and the morbidity associated with advanced disease. The key to achieving the greatest potential effects from screening is providing early access to effective diagnostic imaging and biopsy as well as improved treatment options.

The Solution*

Zebra’s mammography solution works on 2D mammography screening data (FFDM) to identify Regions of Interest (ROI) which are suspicious for malignancy, including but not limited to masses and calcifications.

As published on the Journal of Digital Imaging, volume 30, issue 4


Customers with a mammography screening program and others use the Zebra mammography solution to be notified on suspicious lesions on screening studies and attend those in a timely manner. Using this solution, radiologists who are going through massive amounts of screening data enjoy a “safety net” which helps decrease the amount of non identified suspicious lesions.

How it works

Zebra’s mammography solution uses state of the art machine learning technologies that automatically analyse images for suspicious lesions as they come into the PACS system. Once sent to PACS, Zebra analyzes the images for suspicious lesions fully automatically. Zebra’s insights are then sent back to the PACS using DICOM SR and are seamlessly integrated as annotations on the study images where the algorithm detected suspicious abnormalities.

Mammo-Cad flow

The Value

Out of 40M mammograms performed annually in the US alone, about 10% (sometimes up to 15%) are mistakenly suspected to be cancerous (i.e. false alarm, “False Positives”). This represents 4M or more women having to go through extra examinations, while causing additional costs for the healthcare sector. Solutions that could improve readers accuracy, even by a small percentage, can provide a huge impact on many lives and cost saving to many healthcare institutes.

Zebra Medical Vision is HIPPA and GDPR compliant and has the organizational and technical measures in place to fulfil its fundamental requirements.
Zebra’s AI analysis is performed on de-identified studies, while keeping personal health information (PHI) safe and private.
Read more about our privacy policy.


Interested in learning more? Contact Us