A radiopathomics model for predicting large-number cervical lymph node metastasis in clinical N0 papillary thyroid carcinoma

ObjectivesThis study aimed to develop a multimodal radiopathomics model utilising preoperative ultrasound (US) and fine-needle aspiration cytology (FNAC) to predict large-number cervical lymph node metastasis (CLNM) in patients with clinically lymph node-negative (cN0) papillary thyroid carcinoma (PTC).Materials and methodsThis multicentre retrospective study included patients with PTC between October 2017 and June 2024 across seven institutions. Patients were categorised based on the presence or absence of large-number CLNM in training, validation, and external testing cohorts. A clinical model was developed based on the maximum diameter of thyroid nodules. Radiomics features were extracted from US images and pathomics features were extracted from FNAC images. Feature selection was performed using univariate analysis, correlation analysis, and least absolute shrinkage and selection operator regression. Six machine learning (ML) algorithms were employed to construct radiomics, pathomics, and radiopathomics models. Predictive performance was assessed using the area under the curve (AUC), and decision curve analysis (DCA).ResultsA total of 426 patients with PTC (41.65 +/- 12.47 years; 124 men) were included in this study, with 213 (50%) exhibiting large-number CLNM. The multimodal radiopathomics model demonstrated excellent predictive capabilities with AUCs 0.921, 0.873, 0.903; accuracies (ACCs) 0.852, 0.800, 0.833; sensitivities (SENs) 0.876, 0.867, 0.857; specificities (SPEs) 0.829, 0.733, 0.810, for the training, validation, and testing cohorts, respectively. It significantly outperformed the clinical model (AUCs 0.730, 0.698, 0.630; ACCs 0.690, 0.656, 0.627; SENs 0.686, 0.378, 0.556; SPEs 0.695, 0.933, 0.698), the radiomics model (AUCs 0.819, 0.762, 0.783; ACCs 0.752, 0.722, 0.738; SENs 0.657, 0.844, 0.937; SPEs 0.848, 0.600, 0.540), and the pathomics model (AUCs 0.882, 0.786, 0.800; ACCs 0.829, 0.756, 0.786; SENs 0.819, 0.889, 0.857; SPEs 0.838, 0.633, 0.714).ConclusionThe multimodal radiopathomics model demonstrated significant advantages in the preoperative prediction of large-number CLNM in patients with cN0 PTC.Key PointsQuestionAccurate preoperative prediction of large-number CLNM in PTC patients can guide treatment plans, but single-modality diagnostic performance remains limited.FindingsThe radiopathomics model utilising preoperative US and FNAC images effectively predicted large-number CLNM in both validation and testing cohorts, outperforming single predictive models.Clinical relevanceOur study proposes a multimodal radiopathomics model based on preoperative US and FNAC images, which can effectively predict large-number CLNM in PTC preoperatively and guide clinicians in treatment planning.Key PointsQuestionAccurate preoperative prediction of large-number CLNM in PTC patients can guide treatment plans, but single-modality diagnostic performance remains limited.FindingsThe radiopathomics model utilising preoperative US and FNAC images effectively predicted large-number CLNM in both validation and testing cohorts, outperforming single predictive models.Clinical relevanceOur study proposes a multimodal radiopathomics model based on preoperative US and FNAC images, which can effectively predict large-number CLNM in PTC preoperatively and guide clinicians in treatment planning.Key PointsQuestionAccurate preoperative prediction of large-number CLNM in PTC patients can guide treatment plans, but single-modality diagnostic performance remains limited. FindingsThe radiopathomics model utilising preoperative US and FNAC images effectively predicted large-number CLNM in both validation and testing cohorts, outperforming single predictive models.Clinical relevanceOur study proposes a multimodal radiopathomics model based on preoperative US and FNAC images, which can effectively predict large-number CLNM in PTC preoperatively and guide clinicians in treatment planning.