Dc 65

All patients underwent US examination using EPIQ5 (Philips Ultrasound, Inc., Bothell, Washington, USA; L12-5 linear-array transducer probe, 5 ~ 12 MHz), Aixplorer ultrasound scanner (SuperSonic Imagine, France; SL15-4 linear-array transducer probe, 4 ~ 15 MHz), and Resona 7, Resona 8, Resona R9, Resona I9, and DC-65 (Mindray, Shenzhen, China; L14-5, L14-3, and L13-3 linear-array transducer probes, 3 ~ 14 MHz). The image settings for each examination, including time gain compensation, focus position and dynamic range, were optimized according to the manufacturer’s recommendations.
After physical examination, the whole breast was systematically examined. All cases were evaluated in the supine position, with the upper arm abducted and the bilateral breast fully exposed at the same time. The characteristics of sonogram were observed by routine ultrasonic scanning. The US images were obtained in both longitudinal and transverse sections.

(a) BI-RADS 6 in the US, DR (Digital mammography), or MRI (magnetic resonance imaging). (b) BI-RADS 1 in the US. (c) History of breast surgery. (d) Single-frequency US video data. (e) Missing pathological results. (f) Poor image quality.
In this study, a total of 668 breast tumors (260 malignancy and 408 benign) of US videos from 167 female patients were included and divided into 2 testing sets: (1) In testing set 1, breast US videos were obtained from Resona I9 (Mindray, China) with L14 (frequency range: 3.0-14.0 MHz, central frequency 9 MHz), L9 (frequency range: 2.5-9.0 MHz, central frequency 6.5 MHz) linear array probe. (2) Testing set 2, the data were obtained from DC-65 (Mindray, China) with L13 (frequency range: 3.6-13.5 MHz, central frequency 8 MHz), and L7 (frequency range: 3.0-7.0 MHz, central frequency 4.0 MHz) linear array probe, which was aimed to further evaluate our theory and discoveries.