Objective:

To explore the application value of ultrasound characteristics in predicting the biological risk of gastric gastrointestinal stromal tumors.

Methods:

The ultrasonic characteristics and clinical data of 148 cases of gastric gastrointestinal stromal tumors confirmed by pathology were retrospectively analyzed. Multivariate logi

stic regression analysis was used to screen out the independent influencing factors for the biological risk classification

and the regression equation was established. Receiver operating characteristic (ROC) curve was drawn to analyze the predictive value of logistic regression model.

Results:

There were statistically significant differences in the maximum diameter

echo

cystic changes

morphology

boundary and blood flow signals between the low-risk and medium-high risk gastric gastrointestinal stromal tumor groups in ultrasonic signs（

P

＜0.001）. There were no significant differences in tumor growth site

age of onset

gender and symptom between the two groups （

P

＞0.05）. Multivariate logistic regression analysis showed that the maximum size

echo and morphology of the tumor were independent factors for the identification of gastric gastrointestinal stromal tumors with different risk levels (OR were 1.34

3.81

2.64

and P values were 0.03

0.02

0.04

respectively). The constructed logistic regression equation is logistic (

P

) =-3.03+0.30maximum diameter+1.34 echo+0.97morphology

with the prediction probability

P

=0.50 as the threshold. The sensitivity

specificity

accuracy and area under ROC curve were 78.21%

77.14%

77.70% and 0.89 respectively.

Conclusion:

Logistic regression model based on the maximum size

echo and morphology of tumor body has potential clinical value in predicting the biological risk of gastric gastrointestinal stromal tumor.