Radiological diagnosis of breast cancer: AI for better patient care

Radiological diagnosis of breast cancer: AI for better patient care


Called Mammography Technical Evaluation (MTE), this new feature will help radiologists and radiographers correctly position patients during mammography examinations. It was developed by Hera-MI, the French inventor of Breast-SlimView®, the only global breast cancer screening aid solution based on negativation AI.

Technically insufficient shots

The objective is to limit the dose of x-rays and to reduce the number of technically insufficient shots. Indeed, in the context of organized breast cancer screening, technically insufficient X-rays require a third reading or a recall of the patient, generating enormous stress and additional costs. This innovation will be presented at the Journées francophones de radiologie in Paris in October.

In France, one of the specificities of the national organized breast cancer screening program is the practice of a second reading of all mammograms in the absence of any suspicious abnormality on the first reading.

Santé Publique France reports that in 2017-2018, of the 40,120 breast cancers detected, 2,030 (5.3%) were detected at second reading.

In the event of a technically insufficient X-ray (blurry X-rays, lacunar compression, insufficiently exposed mammary gland), the second reader must request that the X-rays be redone, in order to interpret the mammograms under better technical conditions.

Since 2006, the law indicates “ that before making this decision, he will weigh up a possible loss of opportunity for the woman linked to the insufficient quality of the images against the deleterious effects of a new summons ” and that “ these cases must be few in number (< 1% of screenings carried out), because they generate anxiety and loss of confidence. The decision to reconvene a woman due to a technical insufficiency can be the subject of a consensus of second readers whenever possible. »

Mammography Technical Evaluation

Mammography Technical Evaluation is an innovative feature of the Breast-SlimView® solution on the market. Its algorithm evaluates the quality of images in mammography and tomosynthesis, in order to quickly identify technically insufficient images.

Hera-MI’s AI automatically analyzes several essential criteria such as breast compression, positioning (pectoral and nipple, for example) and finally, the general quality of the image on criteria such as artifacts, folds…

Mammography Technical Evaluation features a personalized tracking and areas for improvement interface that guides radiographers toward quick and effective correction.

Thanks to Breast-SlimView® Mammography Technical Evaluation, radiographers can self-assess their practice, analyze their areas of improvement, in order to standardize good practices. The radiologist, for his part, benefits from qualitative images allowing him to optimize his screening and make his diagnoses more reliable.

“Our AI has real added value in public health. Positioning the patient well and reducing the number of technically insufficient X-rays are challenges in breast cancer screening. Our research will help healthcare professionals — radiographers and radiologists — in their practice. They will also benefit patients, who will avoid a new examination and a lower dose of X-rays. , explains Sylvie DAVILA, President of Hera-MI.

*Negativation: The AI automatically identifies and masks all physiologically normal elements of the breast (vessels, connective tissues, mammary gland, muscles, etc.). Only areas potentially at risk of malignancy are visible on the image. This process reduces the volume of unnecessary information. It allows radiologists to focus on suspicious areas and discover an area at high risk of malignancy on the first reading.