Skip to main content

Table 3 The used 13 features for comparison

From: Comparison of six machine learning methods for differentiating benign and malignant thyroid nodules using ultrasonographic characteristics

Variable

Name/Description

X1

Gender/Sex

X2

irregular margins

X3

ill-defined borders

X4

taller-than-wide shape

X5

hypoechogenicity or marked hypoechogenicity

X6

microcalcification

X7

posterior echo attenuation

X8

a peripheral acoustic halo

X9

an interrupted thyroid capsule

X10

central vascularity

X11

suspected cervical lymph node metastasis

X12

strain elastosonography

X13

contrast-enhanced ultrasound