This study demonstrated morphological characteristics of MRI can hardly diagnose the expression of E-cadherin and Vimentin in rectal cancer. ADC value (Ultra-high b-Value) was positively and negatively correlated with E-cadherin and Vimentin expression. Ktrans and Kep values were negatively and positively correlated with E-cadherin and Vimentin expression. ADC, Ktrans and Kep values had significant diagnostic efficiencies for low E-cadherin and high Vimentin expression. Radiomics signatures with machine learning have excellent diagnostic efficacy for E-cadherin and Vimentin expression, building the model which combined MRI quantitative parameters and radiomics features will improve the predictive performance for E-cadherin and Vimentin expression.
This abstract and the presentation materials are available to members only; a login is required.