Consolidate deep learning, innovation of ultra-tiny object detection, and multi-view attention fusion technology into an algorithm model. The system can identify micro-calcifications, masses, asymmetric shadows, and structural distortions. It also detects concomitant signs such as nipple depression and lymph node enlargement.Referring to the fifth edition of the ACR BI-RADS criteria provides diagnostic recommendations for the BI-RADS classification, and generates structured reports, so it can improve radiologists' diagnostic efficiency and quality.

The system detects automatically acute hemorrhagic and ischemic lesions in head CT/MRI. It can position accurately the location of the lesion, identify the hematoma's classification and measure the volume. At the same time， it can provide a follow-up comparison to assist radiologists to evaluate the state of a stroke.