Aplio 500

Different types of ultrasound diagnostic equipment (LOGIQ E9, Siemens Acuson S2000, Toshiba Aplio 500 and Philips EPIQ 5) were used at the above 2 hospitals, employing a high-frequency linear probe with radial, transverse, and longitudinal scans on both breasts. Ultrasound images were exported with the format of digital imaging and communication in medicine from the picture archiving and communication system database.
In this study, sonographer 1 was responsible for collecting the ultrasound image information of the patients. Preprocessing was carried out using resampling technique, resampling all of the ultrasound images so that they were 1 mm × 1 mm × 1 mm to obviate the disturbances due to the nonuniform spatial resolution. Next, ITK-SNAP software (open source software; http://www.itk-snap.org) was employed to manually outline region of interest that covered the largest cross-sectional area of each breast lesion in the transverse plane. This was carried out independently by sonographer 2 (sonographer 1 from Hospital 2, with more than 5 years’ experience in ultrasonic diagnosis) not knowing a patient's histopathological grade result.

The study consisted of three arms: A, B, and C (Figure 1). Arm A tested four ultrasound devices: the Siemens Acuson S3000, Toshiba Aplio 500 (software version 5.0), Philips Epiq 7, and GE Logiq E9. We chose the Siemens S3000 ultrasound scanner as the reference device because the largest number of studies exist for this ultrasound scanner or its predecessor, the Siemens S2000[27 (link)-29 (link)]. The Siemens S3000 and Philips Epiq 7 devices use p-SWE technology, and the Aplio 500 Toshiba and GE Logiq E9 devices use 2D-SWE technology. Study arm A showed that the Toshiba Aplio 500 device (software version 5.0) generated strongly deviating results compared to the other ultrasound devices tested. Due to the divergent results between Toshiba Aplio 500 (version 5.0) and the other tested devices, especially the reference device, the Toshiba Aplio 500 was tested using software version 6.0 against the Siemens Acuson S3000 in study arm B. In study arm C, the results of study arms A and B were compared to investigate the differences between the two different software versions of the Toshiba Aplio 500. Study arm A was conducted from May 2015 to September 2015 and study arm B from November 2016 to April 2017.

GE Logiq E8, Toshiba Aplio 500, and Philips IU22 ultrasound diagnostic systems and high-frequency linear array probes (5–12 MHz) were used to assess the number, size, echogenicity, border, margin, shape, microcalcification, blood flow, and capsular invasion of the tumor. If the patient had multiple lesions, the largest one was selected as the observation object. The tumors were divided into three groups: the largest diameter <10 mm, 10–20 mm, and >20 mm. The blood flow of the tumor was graded by the Adler semiquantitative method,29 (link) according to the following criteria: Grade I: no obvious blood flow, indicating the absence of obvious blood flow signal inside the tumor; Grade II: small amount of blood flow, showing 1–2 punctate or short rod blood flow signals in the tumor; Grade III: moderate blood flow, showing 3–4 punctate blood flow signals or one long strip of blood flow signal in the tumor; and Grade IV: abundant blood flow, showing >5 punctate blood flow signals, >2 long strip blood flow signals, reticular blood flow signals, or dendritic blood flow signals. At ultrasound, when it was observed that the tumor was in contact with >25% of the adjacent thyroid capsule, or the thyroid capsule was disrupted by the tumor, or the soft tissue around the thyroid gland was invaded, it was defined as capsular invasion.6 (link),30 (link),31 (link)