Of 119 cases of breast tumor enrolled in a comparison report of breast ultrasound by conventional breast ultrasound and AI at Medic Center; the authors, between 2 methods, declared AI helps exactly diagnosing the classified BI-RADS 4c cases while the results of both 2 methods had statistically the same sensibility for all classified BI-RADS. It’s maybe, in our knowledge, the first report concerning AI on breast ultrasound in Vietnam and other southern Asian countries.
But conventional breast ultrasound described more characteristics of the malignancy of the breast tumor than AI that only based on the 5 classic criterii : boundary, circumscribed, axe, echogeneicity, and tumoral form. However AI could use like a screening tool of ultrasound for the malignant breast tumors. In the protocol, biopsy and, in particular conditions, breast MRI are gold standard for correct evaluation of breast tumor.
In condition, AI improved the malignancy for cases of classified BI-RADS 4c and over. And the authors said that naturally needs more informed reports of other studies with more tumor breast cases.