To develop and externally validate an MRI-based deep learning framework for automated 3D segmentation of neck lymph nodes (LNs) in head and neck squamous cell carcinoma (HNSCC), and to assess segmentation accuracy, volumetric agreement, and time efficiency. Axial head and neck MRI data were retrospectively collected from two centers. Each LN was manually delineated following two strategies: margin-excluding approach (excluding a 1–2 mm perinodal margin) and boundary-adherent approach (following the LN contour). For each strategy, nnU-Net (a DL framework) was applied for training three single-sequence models using contrast-enhanced T1-weighted (CE-T1w), T1-weighted (T1w), and T2-weighted (T2w) imaging. The model performance was evaluated in the internal validation cohort (n = 99) and the external test cohort (n = 150). Efficiency was assessed by comparing manual vs. DL-assisted contouring times in a separate cohort (n = 851). In the internal validation cohort, the model trained on CE-T1w images with boundary-adherent annotations achieved the highest overlap (median Dice similarity coefficient of 0.795). CE-T1w model showed high volumetric agreement with manual segmentation (concordance correlation coefficient of 0.939 for margin-excluding and 0.899 for boundary-adherent strategies), and its performance was confirmed in the external test cohort. DL assistance reduced contouring time by 33.14

BACKGROUND:Preoperative discrimination between orbital B-cell lymphoma and inflammatory lesions remains a significant challenge using conventional imaging. This study evaluates the potential of time-dependent diffusion MRI (td-dMRI) alongside diffusion-weighted imaging (DWI) to improve differential diagnosis. METHODS:Patients with suspected orbital tumors were prospectively enrolled between October 2023 and November 2024. All participants underwent td-dMRI using oscillating gradient spin-echo (OGSE) and pulsed gradient spin-echo (PGSE) sequences on a 3 T scanner. Microstructural parameters-including cell diameter (d), cellularity, extracellular diffusivity (Dex), and intracellular volume fraction (Vin)-were derived. Correlations between apparent diffusion coefficient (ADC), d, cellularity, and histopathological metrics were assessed through quantitative morphometric analysis. Diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. RESULTS:Forty-eight patients were enrolled (20 orbital MALT lymphoma, 28 inflammatory lesions). All parameters showed excellent interobserver agreement (ICCs: 0.82-0.93). Strong correlations were observed between histological cell diameter and d (r = 0.75), and between histological cellularity and both td-dMRI-derived cellularity (r = 0.71) and ADC (r = -0.74) (all p < 0.001). Lymphomas exhibited significantly lower ADC, d, and Dex, and higher Vin and cellularity compared to inflammatory lesions (all p < 0.01). Cellularity demonstrated the highest discriminative power (AUC = 0.87), followed by Vin (AUC = 0.82), while ADC showed moderate performance (AUC = 0.74). No significant differences in diagnostic efficacy were observed among the parameters. CONCLUSIONS:The td-dMRI provides highly reproducible, histologically correlated biomarkers that effectively differentiate orbital lymphoma from inflammatory lesions, thereby enabling orbital lesion risk-stratification and providing valuable non-invasive characterization to complement conventional ADC-based assessment for preoperative diagnosis of orbital lesions.

This study aims to develop a multimodal nomogram to predict neoadjuvant chemoimmunotherapy (NCIT) outcomes in head and neck squamous cell carcinoma (HNSCC). Treatment-naive HNSCC patients receiving neoadjuvant NCIT were retrospectively analyzed. Clinical information, conventional MR imaging features, dynamic contrast-enhanced-MRI (DCE-MRI) parameters and ADC values were analyzed in relation to pathological complete response (pCR). The predictive accuracy of clinical and MRI parameters was evaluated using the receiver operating characteristic (ROC) curve, with the area under the curve (AUC) serving as a key metric. Following NCIT, 55.0

To develop and externally validate machine learning radiomic models in a two-center retrospective study based on multiparametric MRI, including apparent diffusion coefficient (ADC) mapping, for distinguishing benign from malignant sinonasal tumors. This retrospective study enrolled 497 patients with pathologically confirmed sinonasal tumors from two centers (Center 1, n = 318; Center 2, n = 179). Data from Center 1 were randomly divided into training (70

Pretreatment determination of histological differentiation grade is critical for prognostic evaluation in laryngeal and hypopharyngeal squamous cell carcinoma (LHSCC). This study aimed to develop a contrast-enhanced CT (CECT)-based Vision Transformer (ViT) model for noninvasive evaluation of histological grades in LHSCC. In this retrospective multicenter study, a total of 1,648 LHSCC patients who underwent CECT scans were enrolled from three hospitals in this study. Participants were divided into a training cohort (n = 1,239), an internal validation cohort (n = 310) from one hospital, and an external validation cohort (n = 99) from the other two hospitals. The diagnostic model integrates a pre-trained ViT for CECT feature extraction and an XGBoost classifier for prediction. The model’s predictive performance was evaluated using the area under the curve (AUC), decision curve analysis (DCA), and calibration curve. The ViT model achieved AUCs of 0.887 (95

PURPOSE:This study aimed to develop a multisequence MRI-based volumetric histogram metrics model for predicting pathological complete response (pCR) in advanced head and neck squamous cell carcinoma (HNSCC) patients undergoing neoadjuvant chemo-immunotherapy (NCIT) and compare its predictive performance with AJCC staging and RECIST 1.1 criteria.METHODS:Twenty-four patients with locally advanced HNSCC from a prospective phase II trial were enrolled for analysis. All patients underwent pre- and post-NCIT MRI examinations from which whole-tumor histogram features were extracted, including T1WI, T2WI, enhanced T1WI (T1Gd), diffusion-weighted imaging (DWI) sequences, and their corresponding apparent diffusion coefficient (ADC) maps. The pathological results divided the patients into pathological complete response (pCR) and non-pCR (N-pCR) groups. Delta features were calculated as the percentage change in histogram features from pre- to post-treatment. After data reduction and feature selection, logistic regression was used to build prediction models. ROC analysis was performed to assess the diagnostic performance.RESULTS:Eleven of 24 patients achieved pCR. Pre_T2_original_firstorder_Minimum, Post_ADC_original_firstorder_MeanAbsoluteDeviation, and Delta_T1Gd_original_firstorder_Skewness were associated with achieving pCR after NCIT. The Combined_Model demonstrated the best predictive performance (AUC 0.95), outperforming AJCC staging (AUC 0.52) and RECIST 1.1 (AUC 0.72). The Pre_Model (AUC 0.83) or Post-Model (AUC 0.83) had a better predictive ability than AJCC staging.CONCLUSION:Multisequence MRI-based volumetric histogram analysis can non-invasively predict the pCR status of HNSCC patients undergoing NCIT. The use of histogram features extracted from pre- and post-treatment MRI exhibits promising predictive performance and offers a novel quantitative assessment method for evaluating pCR in HNSCC patients receiving NCIT.

The purpose of this study was to evaluate the additional value of dynamic contrast-enhanced (DCE) MRI and diffusion weighted MRI (DWI) in differentiation between inflammatory myofibroblastic tumor (IMT) and squamous cell carcinoma (SCC) in the sinonasal cavity. Patients with pathologically proven IMT and SCC in the sinonasal region were enrolled in this retrospective study. All participants underwent conventional MRI and dynamic contrast-enhanced MRI, while a subset of them performed DWI. All the MRI parameters were independently analyzed by two investigators. This retrospective study included 21 patients with IMT and 55 patients with SCC. Significant differences were found in the conventional MR imaging features including mass margin, T2 signal intensity and track sign of maxillary (p < 0.05). For DCE-MRI features, significant differences were found in progressive centripetal continual enhancement and CImax (p < 0.001 and p = 0.026, respectively). A marginal significant difference was found in ADC values between IMT (0.86 ± 0.59) and SCC (1.14 ± 0.25) (p = 0.061). The conventional MRI analysis revealed that the combination of mass margin and track sign of maxillary yielded an accuracy of 81.6

Objective:This study aimed to investigate the application value of fractional bolus injection of a contrast agent combined with double-layer detector spectral CT 50 keV virtual monoenergetic imaging(VMI)in combined imaging of computed tomography urography(CTU)and aortic CT angiography(CTA).Methods:The experimental group included 32 patients who underwent spectral CTU with a Philips IQon dual-layer detector at Beijing Tongren Hospital,Capital Medical University,between March and April 2024.Using fractional bolus injection of a contrast agent and bolus tracking technology,the conventional 120 kVp mixed energy image was reconstructed after scanning for group A1;the 50 keV VMI obtained by spectral reconstruction was group A2.The control group included 32 patients who underwent CTU examination between December 2023 and March 2024 using the same equipment,and the contrast agent split bolus and group injection tracking techniques were used.After scanning,the conventional 120 kVp mixed energy image was reconstructed for group B.The objective evaluation included the CT value,contrast-to-noise ratio(CNR),and signal-to-noise ratio(SNR)of the abdominal aorta and bilateral renal arteries,which were compared among the three groups.The Kruskal-Wallis and Nemenyi tests were used to compare the average CT values of the initial,middle,and lower parts of the ureter.Wilcoxon test was used to compare the effective radiation dose between the experimental and control groups.Subjective evaluation was performed by two senior radiologists on the three groups of CTU and CTA images using the 4-and 3-point methods,respectively.The consistency of the subjective scores of the two radiologists was analyzed using the Kappa test.Results:Among the objective evaluation indices,there were statistically significant differences among groups A2,A1,and B.There was a significant difference in the SNR of the bilateral renal arteries between groups A1 and B.The effective radiation dose of the experimental group was 15.1%lower than that of the control group,which was significantly different.There was no significant difference in the subjective evaluation of the CTA and CTU images among the three groups.The consistency of the CTA and CTU image scoring results between the two physicians was excellent and good,respectively.Conclusion:If CTU examination was performed by fractional injection of a contrast agent,fusion images of the urinary system and aorta were obtained,which not only had a clear anatomical relationship and the same image quality,but also reduced the effective radiation dose.The 50keV virtual monoenergetic image obtained by spectral reconstruction optimizes the image quality of CTU and CTA and has clinical value.

PURPOSE:This study aimed to investigate whether multiparametric magnetic resonance imaging (MRI) including dynamic contrast-enhanced (DCE) and diffusion weighted (DW) MRI can differentiate pleomorphic adenoma (PA) from schwannoma in the parapharyngeal space. METHODS:Forty-six patients with pathologically proven PAs and 47 schwannomas in the parapharyngeal space were enrolled. All patients underwent conventional MRI, and DW-MRI and DCE-MRI were performed in 30 and 33 patients, respectively. Fisher's exact, Mann-Whitney-U tests and Independent samples t-test were used to compare variables between PAs and schwannomas. Multivariate logistic regression analysis was used to examine the diagnostic performance of MRI parameters. RESULTS:The PAs usually show lobulation sign, posterior displacement of ICA and attached to the parotid gland deep leaf, while bird beak configuration, anterior displacement of ICA and involvement of foramen jugular were more commonly seen in the schwannomas(all p < 0.001). The washout rate of PAs was found to be higher than that of schwannomas (p = 0.035), whereas no significance was found in the other DCE-MRI parameters and in ADCs(p > 0.05). Using a combination of conventional MRI features including lobulation sign, bird beak configuration, direction of internal carotid artery(ICA) displacement and attached to the parotid gland in multivariate logistic regression analysis, sensitivity, specificity, and accuracy in differential diagnosis of PAs and schwannomas were 97.8%, 91.5% and 94.6%, respectively. CONCLUSION:Conventional MRI can effectively differentiate PAs from schwannomas in the parapharyngeal space with a high diagnostic accuracy. The DCE-MRI and DWI have limited added diagnostic value to conventional MRI in the differential diagnosis.

Accurate diagnosis and prognosis prediction are conducive to early intervention and improvement of medical care for natural killer/T cell lymphoma (NKTCL). Artificial intelligence (AI)-based systems are developed based on nasopharynx magnetic resonance imaging. The diagnostic systems achieve areas under the curve of 0.905-0.960 in detecting malignant nasopharyngeal lesions and distinguishing NKTCL from nasopharyngeal carcinoma in independent validation datasets. In comparison to human radiologists, the diagnostic systems show higher accuracies than resident radiologists and comparable ones to senior radiologists. The prognostic system shows promising performance in predicting survival outcomes of NKTCL and outperforms several clinical models. For patients with early-stage NKTCL, only the high-risk group benefits from early radiotherapy (hazard ratio = 0.414 vs. late radiotherapy; 95% confidence interval, 0.190-0.900, p = 0.022), while progression-free survival does not differ in the low-risk group. In conclusion, AI-based systems show potential in assisting accurate diagnosis and prognosis prediction and may contribute to therapeutic optimization for NKTCL.

This study aimed to investigate the feasibility of diffusion-weighted imaging (DWI) in combination with conventional MRI features to differentiate sinonasal malignant melanoma (SNMM) from sinonasal squamous cell carcinoma (SNSCC). A total of 37 patients with SNMM and 44 patients with SNSCC were retrospectively reviewed. Conventional MRI features and apparent diffusion coefficients (ADCs) were evaluated independently by two experienced head and neck radiologists. ADCs were obtained from two different regions of interest (ROIs) including maximum slice (MS) and small solid sample (SSS). Multivariate logistic regression analysis was performed to identify significant MR imaging features in discriminating between SNMM and SNSCC. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance. SNMMs were more frequently located in the nasal cavity, with well-defined border, T1 Septate Pattern (T1-SP) and heterogeneous T1 hyperintensity, whereas SNSCCs were more frequently located in the paranasal sinus, with homogenous T1 isointensity, ill-defined border, reticular or linear T2 hyperintensity, and pterygopalatine fossa or orbital involvement (all p < 0.05). The mean ADCs of SNMM (MS ADC, 0.85 × 10−3mm2/s; SSS ADC, 0.69 × 10−3mm2/s) were significantly lower than those of SNSCC (MS ADC, 1.05 × 10−3mm2/s; SSS ADC, 0.82 × 10−3mm2/s) (p < 0.05). With a combination of location, T1 signal intensity, reticular or linear T2 hyperintensity, and a cut-off MS ADC of 0.87 × 10−3mm2/s, the sensitivity, specificity, and AUC were 97.3

Topic: 19. Aggressive Non-Hodgkin lymphoma - Clinical Background: Natural killer/T-cell lymphoma (NKTCL) often presents with extra-nodal involvement of nasal region, thus the prognostic system for NKTCL needs to be further optimized. Magnetic resonance imaging (MRI) of nasopharynx is a routine examination performed for newly-diagnosed NKTCL, however, the information in MRI images might be underused. Artificial intelligence (AI) can identify features from microimaging structures in pixel-level, and has shown great potential in assisting clinical decision making. Aims: Based on nasopharynx MRI images and clinical data from nine medical centers from China, we performed a multicenter retrospective study to construct an AI-based prognostic system for NKTCL. Methods: In total, 288 pathologically-proven NKTCL patients with available pretreatment nasopharynx MRI were included, and divided into the training (n=134), internal validation (n=58), and external validation (n=96) datasets. We used segmented, axial, T1-weighted contrast-enhanced MRI images to construct the prognostic system. The regions of interest of NKTCL lesions were manually segmented. Radiomic features were extracted, screened, and the “MRI score” was obtained after inputting the selected features into a Random Survival Forest model based on progression-free survival. The “total score” was obtained after inputting the MRI score and clinical parameters of each patient into the RSF model. The performance of the prognostic systems was then validated in independent datasets, and compared with several clinical used prognostic models by time-dependent area under curve (AUC) and concordance-index (C-index). Results: The overall score achieved satisfactory performance in internal and external validation set. The 3-year time dependent AUC in the internal and external validation set were 0.863 (95% CI, 0.747 - 0.939) and 0.774 (95%CI, 0.678 - 0.853) for PFS, while 0.830 (95% CI, 0.708 - 0.916) and 0.776 (95%CI, 0.679 - 0.854) for OS. In the internal and external validation set, the AUCs of the total score were significantly higher than the international prognostic index (IPI), the Korean prognostic index (KPI), and the prognostic index of natural killer lymphoma (PINK) (all p < 0.05). In the external validation set, comparing to the IPI, KPI, and PINK, the C-indexes of the total score were significantly higher in predicting PFS (C-index [95% CI]: total score, 0.774 [0.711-0.838] vs. IPI, 0.617 [0.483-0.752], p = 0.010; vs. KPI, 0.678 [0.568-0.789], p = 0.027; vs. PINK, 0.649 (0.538-0.759), p = 0.014). However, when comparing to the IPI and PINK in the internal validation set, the total score showed only numerically but not statistically differences (p=0.071 for IPI, and p=0.051 for PINK).Summary/Conclusion: Our findings suggested that the proposed AI systems exhibited potential value in the prognosis prediction for NKTCL, and might serve as a complement to current clinical risk stratification methods. Keywords: Non-Hodgkin’s lymphoma, Artificial intelligence, NK-T cells, Magnetic resonance imaging

Objective: To investigate the diagnostic performance of multiparametric dynamic contrast-enhanced MRI(DCE-MRI) for the differentiation between benign and malignant larcrimal gland epithelial tumors. Methods: The clinical and imaging data of 104 patients with epithelial tumors of the lacrimal gland who underwent orbital MRI scan and met the inclusion criteria in Beijing Tongren Hospital from January 2011 to December 2017 were retrospectively collected, including 48 males and 56 females, aged from 12 to 77 (43±7) years. Sixty-three cases of benign epithelial tumors and 41 cases of malignant epithelial tumors were examined by DCE-MRI. The parameters of semiquantitative analysis including: time to peak enhancement (Tpeak), maximum enhancement ratio (ERmax), Slope, washout ratio (WR) and time-signal intensity curve (TIC) types. The parameters of quantitative analysis including: volume transfer constant (Ktrans), the extravascular extracellular volume fraction (Ve) and rate constant (Kep). Receiver operating characteristic (ROC) curve analysis was performed for DCE-MRI parameters with statistically significant differences, the area under the curve (AUC) was calculated, the diagnostic threshold was determined, and the diagnostic performance was evaluated. Logistic regression analysis was used to determine the best parameters for differential diagnosis of benign and malignant epithelial tumors of the lacrimal gland. Results: For the semiquantitative analysis of DCE-MRI, malignant lacrimal gland epithelial tumor had a significantly shorter Tpeak than benign masses [(103.77±57.87) s vs (187.80±77.01) s,P<0.001)], while had a higher value in ERmax, Slope [M(Q1,Q3)] and WR in malignant masses compared with benign one [1.55±0.39 vs 1.36±0.33; 1.76 (0.97,2.27) vs 0.62 (0.50,0.93); 7.70%(1.40%, 21.60%)% vs 0(0, 0),all P<0.05)].The TICs of benign lacrimal tumors mainly showed a persistent type (49/63),while most malignant lacrimal tumors mainly showed a plateau type (25/41). For the quantitative analysis of DCE-MRI, the values of Ktrans and Kep[M(Q1,Q3)] in malignant tumors were significantly greater than those of benign tumors (0.99±0.52/min vs 0.43±0.23/min, P<0.001; 1.33(0.83, 1.55)/min vs 0.55(0.46, 0.68)/min, P<0.001). No significant difference in Ve was found between the groups (0.76±0.20 vs 0.73±0.22,P=0.467). Through the statistical analysis, TIC types (OR=3.887,95%CI: 1.409-10.725) and Ktrans(OR=50.979,95%CI: 6.046-429.830) can provide superior diagnostic performance for predicting malignant lacrimal gland epithelial tumors, with a sensitivity of 78.05%, specificity of 77.78%,and sensitivity of 70.73%, specificity of 95.24%, respevtively. Furthermore, the comprehensive diagnostic performance of Ktrans in AUC was proven to be significantly better than that of TIC [0.875 (0.796-0.932) vs 0.798 (0.708-0.870),P=0.049]. Conclusions: Multiparametric DCE-MRI is helpful for the differential diagnosis of benign and malignant epithelial tumors of lacrimal gland. TIC type and Ktrans have higher diagnostic value, and the diagnostic performance of Ktrans is better than that of TIC.

PurposePreoperative assessment of extraocular muscle invasion is essential for therapeutic strategies and prognostic evaluation. The aim of this study was to assess the diagnostic accuracy of MRI for evaluation of extraocular muscle (EM) invasion by malignant sinonasal tumors.Materials and methodsConsecutively, 76 patients of sinonasal malignant tumors with orbital invasion were included in the present study. Preoperative MRI imaging features were analyzed by two radiologists independently. The diagnostic performances of MR imaging features for detecting EM involvement were evaluated by comparing imaging findings to histopathology data.ResultsA total of 31 extraocular muscles were involved by sinonasal malignant tumors in 22 patients, including 10 medial rectus muscles (32.2%), 10 inferior rectus muscles (32.2%), 9 superior oblique muscles (29.1%), and 2 external rectus muscles (6.5%). The EM involved by sinonasal malignant tumors usually showed relatively high signal intensity on T2-weighted images, indistinguishable from the tumor, nodular enlargement and abnormal enhancement (p = 0.001, < 0.001, < 0.001 and < 0.001, respectively). Using a combination of EM abnormal enhancement and indistinguishable from the tumor in multivariate logistic regression analysis, sensitivity, specificity, positive predictive value, negative predictive value and diagnostic accuracy for detecting orbital EM invasion by sinonasal tumors were 93.5, 85.2, 76.3, 96.3 and 88%, respectively.ConclusionMRI imaging features show high diagnostic performance for the diagnosis of extraocular muscle invasion by malignant sinonasal tumors.

Post‐laminar optic nerve invasion (PLONI) is a high‐risk factor for the metastasis of retinoblastoma (RB). Unlike conventional MRI, diffusion‐weighted imaging (DWI) reflects histopathological features, and may aid the assessment of PLONI.

To investigate the value of MRI-based radiomic features integrated with clinical indicators for survival prediction in patients with extranodal natural killer/T-cell lymphoma, nasal-type (ENKTL). One-hundred and sixty-five patients with ENKTL who underwent pretreatment MRI were enrolled. Patients were randomly divided into training (n = 115) and validation (n = 50) sets. A radiomic signature (R-signature) was generated using the least absolute shrinkage and selection operator regression. Kaplan–Meier analysis and univariate Cox proportional hazards model were used to determine the association of the R-signature and clinical variables with overall survival (OS) and progression-free survival (PFS). Clinical models and combined clinical-R-signature models were constructed by multivariable Cox regression analysis, respectively. The R-signature achieved C-index of 0.666 and 0.684 (training set) and 0.679 and 0.691 (test set) for the prediction of OS and PFS, respectively. For both OS and PFS prediction, the C-index was comparable between the R-signature and clinical model both in the training cohort (OS: C-index = 0.666 vs. 0.719, p = 0.284; PFS: C-index = 0.684 vs. 0.725, p = 0.439) and the validation cohort (OS: C-index = 0.679 vs. 0.665, p = 0.878; PFS: C-index = 0.691vs.0.668, p = 0.803), respectively. The combined clinical-R-signature models achieved better predictive performance than the R-signature in the training cohort (OS: C-index = 0.741vs.0.666, p = 0.032; PFS: C-index = 0.762 vs. 0.684 p = 0.020), respectively. The differences did not reach statistical significance in the validation cohort (p > 0.2). The radiomic signature extracted from baseline MRI can predict outcomes of patients with ENKTL, and the combination of MRI radiomic signature and clinical predictors may further improve the predictive performance in patients with ENKTL.

Objective:To investigate the MRI features of the primary sinonasal malignant melanoma (SMM) and evaluate the signal pattern based on T 1WI and T 2WI, in order to improve the diagnostic accuracy of SMM. Methods:The MRI findings of 63 SMM cases confirmed by pathology from April 2007 to November 2018 at Beijing Tongren Hospital, Capital Medical University were analyzed retrospectively. The signal intensity of malignant melanoma was classified into four types(Ⅰ—Ⅳ) according to the proportion of signal areas of the largest slice of the tumor on T 1WI and T 2WI. The classification criteria according to T 1WI: type Ⅰ, the area of hyperintensity was ≥50%; type Ⅱ, the area of hyperintensity was <50%; type Ⅲ, the tumor did not show hyperintensity, and the area of isointensity was ≥50%; type Ⅳ, the tumor did not have high signal area, and the area of low signal was ≥50%. The classification criteria according to T 2WI: type Ⅰ, the area of low signal in the tumor was ≥50%; type Ⅱ, the area of low signal was <50%; type Ⅲ, the tumor did not contain low signal area, and the area of isointensity was ≥50%; type Ⅳ, the tumor did not have low signal area, and the area of high signal intensity was ≥50%. The proportion of each type was calculated. Results:According to T 1WI, typeⅠwas identified in 27 cases (42.9%, 27/63), typeⅡ in 25 cases (39.7%, 25/63), type Ⅲ in 4 cases (6.3%, 4/63), and type Ⅳ in 7 cases (11.1%, 7/63). According to T 2WI, type Ⅰwas demonstrated in 29 cases (46.0%, 29/63), type Ⅱ in 28 cases (44.4%, 28/63), type Ⅲ in 2 cases (3.3%, 2/63), and type Ⅳ in 4 cases (6.3%, 4/63). There were 16 cases classified as type I based on T 1WI and T 2WI. Conclusions:Typical and atypical SMM can be identified according to signal patterns. The typeⅠsignal pattern of SMM cases on T 1WI and T 2WI is typical and can be easily diagnosed, but the proportion was less than 50%. For atypical SMM, malignant melanoma should be strongly suspected if hyperintense on T 1WI or hypointense on T 2WI is found.

Purpose To develop and validate an MRI-based radiomics model in differentiation between sinonasal primary lymphomas and squamous cell carcinomas (SCCs). Materials and methods One-hundred-and-fifty-four patients were enrolled (74 individuals with SCCs and 80 with lymphomas). After feature analysis and feature selection with variance threshold and least absolute shrinkage and selection operator (LASSO) methods, an MRI-based radiomics model with the support vector machine (SVM) classifier was constructed in differentiation between lymphomas and SCCs. Areas under the receiver operating characteristic curves (AUCs) of the MRI-based radiomics model were compared with those of radiologists using Delong test. Results Five features (T1 original shape Compactness2, T1 wavelet-HHH first-order Total Energy, T2 wavelet-HLH GLCM Informational Measure of Correlation1, T1 wavelet-LHL GLCM Inverse Variance and T1 square GLRLM Long Run Low Gray Level Emphasis) were finally selected in the radiomics model. The AUC values in differentiation between lymphomas and SCCs were 0.94 for the training dataset and 0.85 for the validation dataset, respectively. For all the patient datasets, the AUC values of radiomics model, readers 1, 2 and 3 were 0.92, 0.76, 0.77 and 0.80, respectively. For the validation datasets, no significant difference was found between the AUCs of the radiomics model and those of the three radiologist (P = 0.459, 0.469, 0.738 for radiologist 1, 2 and 3, respectively). Conclusion An MRI-based radiomics model can help to differentiate sinonasal lymphomas from SCCs with high accuracy.

目的:探讨婴幼儿在非镇静状态下采用16cm宽体探测器行颞骨CT扫描的可行性.方法:选取154例婴幼儿(0～6岁)采用16cm宽体探测器行颞骨CT扫描,分为A和B组,A组为77例在镇静状态下行CT扫描的对照组,B组为77例在非镇静状态下行CT扫描的实验组,两组扫描及重建参数相同,重组横轴面、冠状面图像,测量和计算对比噪声比(CNR),由两位放射医师对图像进行主观评分,并比较分析两组的CNR和图像主观评分有无差异,符合正态分布的数据采用独立样本t检验方法,不符合的采用秩和检验方法.结果:两名医师主观评分一致性良好(Kappa值为0.784).两组主观评分0～1.0岁分别为62.13±4.07和62.35±4.01,1.1～3.0岁为62.15±2.99和63.08±2.32,3.1～6.0岁为62.93±2.89和62.34±4.46,两组之间差异均无统计学意义(P>0.05).两组0～1.0岁横轴面图像CNR分别为14.18±0.44和14.36±0.65,1.1～3.0岁分别为14.24±0.65和13.94±0.46,3.1～6.0岁分别为13.96±0.33和13.96±0.49;两组0～1.0岁冠状面图像的CNR分别为23.43±1.02和23.20±1.20,1.1～3.0岁分别为23.07±1.23和22.77±0.920,3.1～6.0岁为22.59±1.02和22.67±1.06,两组之间差异均无统计学意义(P>0.05).镇静组阳性率85.71％,非镇静组阳性率88.31％.结论:婴幼儿在非镇静状态下使用16厘米宽体探测器行颞骨CT扫描是可行的,其图像质量与镇静状态下图像质量没有差异.

Purpose To evaluate whether imaging features on conventional magnetic resonance imaging (MRI) can differentiate sinonasal extranodal natural killer/T cell lymphomas (ENKTL) from diffuse large B cell lymphoma (DLBCL). Methods Consecutively, pathology-proven 59 patients with ENKTL and 27 patients with DLBCL in the sinonasal region were included in this study. Imaging features included tumor side, location, margin, pre-contrast T1 and T2 signal intensity and homogeneity, post-contrast enhancement degree and homogeneity, septal enhancement pattern, internal necrosis, mass effect, and adjacent involvements. These imaging features for each ENKTL or DLBCL on total 86 MRI scans were indicated independently by two experienced head and neck radiologists. The MRI-based performance in differential diagnosis of the two types of lymphomas was evaluated by multivariate logistic regression analysis. Results All ENKTLs were located in the nasal cavity, with ill-defined margin, heterogeneous signal intensity, internal necrosis, marked enhancement of solid component on MRI, whereas DLBCLs were more often located in the paranasal sinuses, with MR homogenous intensity, mild enhancement, septal enhancement pattern, and intracranial or orbital involvements (allP < 0.05). Using a combination of location, internal necrosis and septal enhancement pattern of the tumor in multivariate logistic regression analysis, sensitivity, specificity, and accuracy in differential diagnosis of ENKTL and DLBCL were 100%, 79.4%, and 91.9%, respectively, for radiologist 1, and were 98.3%, 81.5%, and 93.0%, respectively, for radiologist 2. Conclusion MRI can effectively differentiate ENKTL from DLBCL in the sinonasal region with a high diagnostic accuracy.