Stages of artificial intelligence development in mammographic image analysis: domestic and foreign experience

Breast cancer is one of the most common cancers in women worldwide. Early diagnosis and screening play a key role in the successful treatment of this disease. Currently, the most common and effective diagnostic method is mammography.
However, the percentage of errors in the interpretation of mammographic images is still high, which leads to a late diagnosis of breast cancer. In this regard, at the end of the XX century, the first developments of auxiliary tools based on artificial intelligence began to improve the efficiency of radiologists.
The first computer aided detections did not meet expectations, but the improvement of the approach to creating artificial intelligence led to the first successes and further development of this field.
This review article discusses modern approaches to creating diagnostic systems based on artificial intelligence, the experience of foreign and Russian developers, and the main AI systems that currently exist, indicating their diagnostic efficiency indicators.