Your browser doesn't support javascript.
loading
Differential diagnosis of inverted papilloma and polyp of nasal cavity and paranasal sinus based on MRI imaging model / 安徽医科大学学报
Article en Zh | WPRIM | ID: wpr-1038585
Biblioteca responsable: WPRO
ABSTRACT
Objective@#To explore the value of T2WI and enhanced T1WI magnetic resonance imaging models in differentiating inverted papilloma and polyp of nasal cavity and paranasal sinus.@*Methods@#54 cases of inverted papilloma (NIP) and 51 cases of polyp ( NP) with complete T2WI and enhanced T1WI images confirmed by pathology were collected.ITK snap was used to outline all levels of the lesion.Pyromics was used to extract image omics features.Firstly,the mRMR) was used for feature extraction,and then rfe-SVM was used to remove the feature of minimum score and establish a prediction model.The sensitivity and specificity of ROC curve were used to evaluate the performance of the model,and were verified in the validation set. @*Results@# A total of 1 133 image omics features were extracted,and 30 features were retained after mrmr dimensionality reduction for the establishment of prediction models.The AUC value of T1WI enhanced prediction model was : training set 0. 98,validation set 0. 95,the sensitivity and specificity of training set were 89. 7% and 100% respectively,and the sensitivity and specificity of validation set were 93. 8% and 93. 3% respectively.The AUC value of T2WI prediction model was : training set 0. 95,validation set 0. 91,the sensitivity and specificity of training set were 82. 1% and 95. 6% respectively,and the sensitivity and specificity of validation set were 93. 8% and 84. 2% respectively.@*Conclusion@#MRI based on T1WI enhancement prediction model and T2WI prediction model have certain value in differentiating inverted papilloma and polyp of nasal cavity and paranasal sinus.
Palabras clave
Texto completo: 1 Índice: WPRIM Idioma: Zh Revista: Acta Universitatis Medicinalis Anhui Año: 2023 Tipo del documento: Article
Texto completo: 1 Índice: WPRIM Idioma: Zh Revista: Acta Universitatis Medicinalis Anhui Año: 2023 Tipo del documento: Article