Deep Learning Radiomics Model Based on Multiparametric MRI to Predict Extrathyroidal Extension in Papillary Thyroid Carcinoma

Abstract

Motivation: Preoperative prediction of extrathyroidal extension could impact the staging and surgical strategy of papillary thyroid carcinoma. Goal(s): Our goal is to establish a DL-combined model to improve prediction performance of extrathyroidal extension. Approach: We constructed a DL radiomics nomogram model based on T2WI, DWI, ADC and delay-phase contrast-enhanced MRI and evaluate the diagnostic performance through area under the receiver operating characteristic curve and decision curve analysis. Results: The combined DL radiomics nomogram predicted ETE with an AUC of 0.936 in training cohort and 0.881 in validation cohort, and the model performed consistently across 1.5T and 3.0T MRI. Impact: This is the first DL radiomics model based on multiparametric MRI for prediction of ETE in PTC, and it could be used as a complement to ultrasound evaluation in clinical practice for PTC patients.

Ähnliche Arbeiten