To examine the prognostic value of a machine learning model trained with pretreatment MRI radiomic features in the assessment of patients with nonmetastatic nasopharyngeal carcinoma (NPC) who are at risk for 3-year disease progression after intensity-modulated radiation therapy and to explain the radiomics features in the model. A total of 277 patients with nonmetastatic NPC admitted between March 2008 and December 2014 at two imaging centers were retrospectively reviewed. Patients were allocated to a discovery or validation cohort based on where they underwent MRI (discovery cohort, n = 217; validation cohort, n = 60). A total of 525 radiomics features extracted from contrast material–enhanced T1- or T2-weighted MRI studies and five clinical features were subjected to radiomic machine learning modeling to predict 3-year disease progression. Feature selection was performed by analyzing robustness to resampling, reproducibility between observers, and redundancy. Features for the final model were selected with Kaplan-Meier analysis and the log-rank test. A support vector machine was used as the classifier for the model. To interpret the pattern learned from the model, Shapley additive explanations (SHAP) was applied. The final model yielded an area under the receiver operating characteristic curve of 0.80 in both the discovery (95% bootstrap confidence interval: 0.80, 0.81) and independent validation (95% bootstrap confidence interval: 0.73, 0.89) cohorts. Analysis with SHAP revealed that tumor shape sphericity, first-order mean absolute deviation, T stage, and overall stage were important factors in 3-year disease progression. These results add to the growing evidence of the role of radiomics in the assessment of NPC. By using explanatory techniques, such as SHAP, the complex interaction of features learned by the model may be understood. © RSNA, 2019 Supplemental material is available for this article.

Background: This study seeks to assess the utility of T1 and T2 mapping in distinguishing metastatic lymph nodes from reactive lymphadenopathy in patients with head and neck squamous cell carcinoma (HNSCC), using diffusion-weighted imaging (DWI) as a comparison. Methods: Between July 2017 and November 2019, 46 HNSCC patients underwent neck MRI inclusive of T1 and T2 mapping and DWI. Quantitative measurements derived from preoperative T1 and T2 mapping and DWI of metastatic and non-metastatic lymph nodes were compared using independent samples t-test or Mann–Whitney U test. Receiver operating characteristic curves and the DeLong test were employed to determine the most effective diagnostic methodology. Results: We examined a total of 122 lymph nodes, 45 (36.9%) of which were metastatic proven by pathology. Mean T2 values for metastatic lymph nodes were significantly lower than those for benign lymph nodes (p < 0.001). Conversely, metastatic lymph nodes exhibited significantly higher apparent diffusion coefficient (ADC) and standard deviation of T1 values (T1_SD) (p < 0.001). T2 generated a significantly higher area under the curve (AUC) of 0.890 (0.826–0.954) compared to T1_SD (0.711 [0.613–0.809]) and ADC (0.660 [0.562–0.758]) (p = 0.007 and p < 0.001). Combining T2, T1_SD, ADC, and lymph node size achieved an AUC of 0.929 (0.875–0.983), which did not significantly enhance diagnostic performance over using T2 alone (p = 0.089). Conclusions: The application of T1 and T2 mapping is feasible in differentiating metastatic from non-metastatic lymph nodes in HNSCC and can improve diagnostic efficacy compared to DWI.

Purpose: To evaluate changes in the levator palpebrae superioris (LPS) muscle on 3.0 T magnetic resonance imaging (MRI) after triamcinolone acetonide injection for treating upper lid retraction (ULR) with Graves’ ophthalmopathy (GO) and to explore the value of LPS muscle quantitative measurement for clinical treatment. Methods: Patients with GO showing ULR were studied retrospectively and they underwent 3.0 T MRI scans before and after subconjunctival injection o f triamcinolone acetonide. The largest thickness (T) and highest signal intensity (SI) of LPS muscle on the affected eyes were measured in the sequences of coronal T2-weighted, fat-suppressed fast spin echo imaging (T2WI-fs) and T1-weighted, fat-suppressed, contrast-enhanced fast spin echo imaging (T1WI-fs + C), respectively. The SI ratio (SIR) (LPS muscle SI/ipsilateral temporalis SI) was calculated individually. Depending on the therapeutic effect, patients were divided into effective group and non-effective group. Independent t-test was used to compare SIR and T of LPS muscle in different treatment groups before treatment, and paired sample t-test was used to compare SIR and T of LPS muscle before and after treatment. Then cut-off level for predicting therapeutic effect and the receiver operating characteristic curve (ROC) curve were analyzed. Results: Sixty-two patients (77 eyes) were enrolled. After treatment, the T of LPS muscle showed significant decrease in all sequences in both effective and non-effective treatment groups. However, changes in SIR of LPS muscle in the two groups were different; SIR of LPS muscle on T2WI-fs and T1WI-fs + C decreased after treatment in the effective group (P _T2 < 0.001, P _{T1 + C} < 0.001) and SIR of LPS muscle showed no statistically difference in all sequences (all P > 0.05) in the non-effective group. There was a correlation between SIR of LPS muscle before treatment and after treatment with triamcinolone acetonide injection, which was that SIR of LPS muscle in the effective treatment group was lower than that in the non-effective treatment group on T1WI-fs + C (P < 0.001). SIR of LPS muscle on T1WI-fs + C showed 87.5% sensitivity and 66.7% specificity to predict therapeutic effect (area under the ROC curve [AUC] = 0.840). Conclusion: In GO patients with ULR, 3.0 T MRI can be used to evaluate the response of triamcinolone acetonide injection. SIR of LPS may be a predictor of its efficacy.

To evaluate the image quality of low-dose temporal bone computed tomography (CT) in otitis media and mastoiditis patients by using deep learning reconstruction (DLR). A total of ninety-seven temporal bones from 53 consecutive adult patients who had suspected otitis media and mastoiditis and underwent temporal bone CT were prospectively enrolled. All patients underwent high resolution CT protocol (group A) and an additional low-dose protocol (group B). In group A, high resolution data were reconstructed by filter back projection (FBP). In group B, low-dose data were reconstructed by DLR mild (B1), DLR standard (B2) and DLR strong (B3). The objective image quality was analyzed by measuring the CT value and image noise on the transverse image and calculating the signal-to-noise ratio (SNR) on incudomallear joint, retroauricular muscle, vestibule and subcutaneous fat. Subjective image quality was analyzed by using a five-point scale to evaluate nine anatomical structures of middle and inner ear. The number of temporal bone lesions which involved in five structures of middle ear were assessed in group A, B1, B2 and B3 images. There were no significant differences in the CT values of the four reconstruction methods at four structures (all p > 0.05). The DLR group B1, B2 and B3 had significantly less image noise and a significantly higher SNR than group A at four structures (all p < 0.001). The group B1 had comparable subjective image quality as group A in nine structures (all p > 0.05), however, the group B3 had lower subjective image quality than group A in modiolus, spiral osseous lamina and stapes (all p < 0.001), the group B2 had lower subjective image quality than group A in modiolus and spiral osseous lamina (both p < 0.05). The number of temporal bone lesions which involved in five structures for group A, B1 and B2 images were no significant difference (all p > 0.05), however, the number of temporal bone lesions which involved in mastoid for group B3 images were significantly more than group A (p < 0.05). The radiation dose of high resolution CT protocol and low-dose protocol were 0.55 mSv and 0.11 mSv, respectively. Compared with high resolution CT protocol, in the low-dose protocol of temporal bone CT, DLR mild and standard could improve the objective image quality, maintain good subjective image quality and satisfy clinical diagnosis of otitis media and mastoiditis patients.