The combination of ultrasound and artificial intelligence is a new branch in the digital medical field, which is also one of the hot spots in the digital medical industry. Ultrasound images contain huge amount of information, and even experienced doctors sometimes struggle to cope with. Artificial intelligence has great potentials to help ultrasound doctors find patterns from ultrasound images then not only to improve testing efficiency, but also reduce the error rate in diagnosis. 


The computer can automatically identify and locate the tumor area from the ultrasonic image, and then evaluate the malignancy degree of tumor. For now, the diseases of interest include thyroid nodules, breast, carotid plaques and liver cancer.

We have established a database containing large number of ultrasound image samples, which can be used to analyze and identify ultrasonic image by trained convolutional neural network. At present, the accuracy of ultrasonic intelligent diagnostic system in industry can be compared with that of clinician, and the sensitivity even better.

Ultrasound intelligent diagnostic system can help doctors’ ultrasound imaging diagnosis, which will not only dramatically save doctors' time,  but also allow patients to get more accurate diagnostic recommendations and better personalized treatments.