Fang et al reported a model for predicting papillary carcinoma in a thyroid nodule based on its ultrasonographic features. The authors are from Nanjing Medical University and affiliated hospitals in Nanjing, China.
Patient selection: thyroid nodule
Parameters:
(1) age in years
(2) echogenicity
(3) shape
(4) margin
(5) border
(6) capsular invasion
(7) length and width
(8) interal composition
(9) calcifications
(10) central vascularity
(11) echogenic hilum in cervical lymph nodes
(12) vascularization of cervical lymph nodes
Parameter
Finding
Points
age in years
< 45 years
0.80
>= 45 years
0
echogenicity
hypo
1.31
other
0
shape
irregular
0.76
regular
0
margin
spiculate
1.19
ill-defined
0.82
other
0
border
indefinite
0.90
other
0
capsular invasion
no
0
yes
2.05
length and width
taller than wide
1.08
wider than tall
0
neither
0
internal composition
solid
0.90
partially cystic
0
cystic
0
calcifications
micro
1.37
mixed micro and macro
1.76
macro only
0.08
neither
0
central vascularity
no
0
yes
0.74
LN echogenic hilum
no
0
yes
1.32
LN vascularization
increased
1.27
not increased
0
value of X =
= SUM(points for all of the parameters) -3.75
probability of papillary thyroid carcinoma =
= 1 / (1 + EXP((-1) * X))
Performance:
• The area under the ROC curve is 0.92.
To read more or access our algorithms and calculators, please log in or register.
