BREAST

The Breast project aims to develop a technology, based on deep learning neural networks, exploiting the power of Artificial Intelligence (AI) for the improvement of mammographic cancer prevention. The AI models on with the technology is based have been developed by the Health Triage Team together with Researchers from the DAUIN (Department of Automatic and Informatics) of the Polytechnic of Turin.

Today 93% of the screening mammograms carried out are negative, a comforting figure, but highly time consuming for the professional resources involved. The goal of Health Triage technology is to overturn the current paradigm, improving its efficiency. The screening mammograms acquired are processed by AI Triage technology, those identified as negative, with 100% absolute certainty, are reprogrammed at the next control exam. This optimizes the work of radiologists so that they can devote their precious time to women who need a more careful reading.

Every year, all women aged 45 to 74 living in Europe are invited and undergo a screening mammography exam. Unfortunately, there is a professional deficit. In several European territories and in some Italian regions there is a lack of doctors to carry out this activity with the right attention and in the necessary dimensions. Health Triage technology can help solve this problem by enabling the achievement of optimal coverage goals set by screening program guidelines. AI Triage technology can offer an adequate response to the growing demand for clinical and diagnostic quality by the population, allowing a wide spread of these prevention initiatives in favor of women, essential for saving lives.

BREAST

The Breast project aims to develop a technology, based on deep learning neural networks, exploiting the power of Artificial Intelligence, for the improvement of mammographic cancer prevention. Today 93% of the screening mammograms carried out are negative, a comforting figure, but highly time consuming for the professional resources involved. The goal of Health Triage technology is to overturn the current paradigm, improving its efficiency. The screening mammograms acquired are processed by AI Triage technology, those identified as negative, with 100% absolute certainty, are reprogrammed at the next control exam. This optimizes the work of radiologists so that they can devote their precious time to women who need a more careful reading.

WHAT ARE THE GOALS OF AI TRIAGE?

Identification of definitely negative mammograms in the screening process (currently negative mammograms are 93% of the exams)
Significant reduction of mammograms to be reported
Improvement of the efficiency of the radiologist up to 143%
Increased precision in clinical diagnosis by eliminating doubtful cases or errors due to fatigue as much as possible
Cost reduction for subsequent diagnostic investigations for public or private health facilities

SPREAD OF THE DISEASE WORLDWIDE

It is the most common cancer in the female population in Western countries
It is the leading cause of cancer death in women in Italy (Source: Airc)
2 million new cases are identified every year worldwide
In Europe, 1 in 10 women are at risk of developing breast cancer
This mortality can be reduced when the disease is diagnosed early before symptoms occur

Throughout Europe there are free screening programs, organized by invitation and supported by awareness campaigns. All screening programs offer the female population, as recommended by the WHO (World Health Organization), a 2D mammography exam. The organizational and logistical commitment is enormous. Just think that in Italy alone two million mammograms are performed per year. The main screening is based on 2D mammography, but a new technique called Tomosynthesis or 3D mammography has also been developing for some years. The invitation to screening programs is stratified according to the age group and takes place via personalized letters and/or by telephone. A level of standardization that has simplified the process of involving the population in prevention programs, but does not adequately take into account the peculiarities of each individual case. In this sense, there is a need to develop personalized screening, based on the clinical and historical characteristics of the patients.