Radiomics Analysis and Correlation With Metabolic Parameters in Nasopharyngeal Carcinoma Based on PET/MR Imaging

Abstract

Objective: Accurate staging is of great importance in treatment selection for patients with nasopharyngeal carcinoma (NPC). The aims of this study were to construct radiomic models of NPC staging based on positron emission tomography (PET) and magnetic resonance (MR) images and to investigate the correlation between metabolic parameters and radiomic features. Methods: A total of 100 consecutive cases of NPC (70 in training and 30 in the testing cohort) with undifferentiated carcinoma confirmed pathologically were recruited. Metabolic parameters of the local lesions of NPC were measured. A total of 396 radiomic features based on PET and MRI images were calculated [including histogram, Haralick, shape factor, gray level co-occurrence matrix (GLCM), and run length matrix (RLM)] and selected [using maximum relevance and minimum redundancy (mRMR) and least shrinkage and selection operator (LASSO)], respectively. The logistic regression models were established according to these features. Finally, the relationship between the metabolic parameters and radiomic features was analyzed. Results: We selected the nine most relevant radiomic features (six from MR images and three from PET images) from local NPC lesions. In the PET model, the area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, and the specificity of the training group were 0.84, 0.75, 0.90, and 0.69, respectively. In the MR model, those metrics were 0.85, 0.83, 0.75, and 0.86, respectively. Pearson's correlation analysis showed that the metabolic parameters had different degrees of correlation with the selected radiomic features. Conclusion: The PET and MR radiomic models were helpful in the diagnosis of NPC staging. There were correlations between the metabolic parameters and radiomic features of primary NPC based on PET/MR. In the future, PET/MR-based radiomic Feng et al. NPC Radiomics Based on PET/MR Imaging models, with further improvement and validation, can be a more useful and economical tool for predicting local invasion and distant metastasis of NPC.

A Systematic Review of PET Textural Analysis and Radiomics in Cancer

“…We collected information from 60 articles that satisfy our selection criteria and evaluated head and neck cancer [ 28 , 82 , 130 , 131 , 132 , 133 , 134 , 135 , 136 , 137 , 138 , 139 , 140 , 141 , 142 , 143 , 144 , 145 , 146 , 147 , 148 , 149 , 150 , 151 , 152 , 153 , 154 , 155 , 156 , 157 , 158 , 159 , 160 , 161 , 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , ...…”

A Comprehensive Review on Radiomics and Deep Learning for Nasopharyngeal Carcinoma Imaging

“…In [ 95 ], 100 consecutive cases of NPC were recruited, and nine of the most relevant radiomic features were selected from features extracted from PET and MRI using LASSO. A predictive model of NPC staging was established based on logistic regression, and the total AUC in PET and MR were 0.84 and 0.85, respectively.…”

Value of PET imaging for radiation therapy

“…Furthermore, PET/CT and PET/MRI data can be used for automated high-throughput radiomics analyses [224,225]. In such studies, standardised quantitative image characteristics are extracted to develop models that support the diagnosis of tumour diseases, the prediction of therapy adaptation, or the prognosis of therapy response, using modern methods of artificial intelligence [226][227][228][229][230][231]. For applicability in clinical practice, efforts on a standardised and reproducible radiomics workflow are decisive [232][233][234][235][236][237].…”