Application of Artificial Intelligence in Breast Imaging

Breast imaging AI technology emerged in the 1960s, initially focusing on advancements in computer-aided detection/diagnosis (CAD) systems. While traditional CAD (conventional AI) systems have been implemented in numerous hospitals throughout China, their capabilities are limited, their performance is subpar, and they exhibit high false-positive rates in lesion detection, leading to a significant performance bottleneck. Consequently, the clinical value of traditional CAD systems has not been fully acknowledged. Nonetheless, the notion of AI aiding radiologists in reducing missed diagnoses and alleviating workload has gained widespread recognition, prompting researchers to explore various perspectives for enhancing the practicality of breast AI systems.