Deep learning for the automatic detection and segmentation of parotid gland tumors on MRI.

OBJECTIVES: Parotid gland tumors (PGTs) often occur as incidental findings on magnetic resonance images (MRI) that may be overlooked. This study aimed to construct and validate a deep learning model to automatically identify parotid glands (PGs) with a PGT from normal PGs, and in those with a PGT to segment the tumor. MATERIALS AND METHODS: The nnUNet combined with a PG-specific post-processing procedure was used to develop the deep learning model trained on T1-weighed images (T1WI) in 311 patients (180 PGs with tumors and 442 normal PGs) and fat-suppressed (FS)-T2WI in 257 patients (125 PGs with tumors and 389 normal PGs), for detecting and segmenting PGTs with five-fold cross-validation. Additional validation set separated by time, comprising T1WI in 34 and FS-T2WI in 41 patients, was used to validate the model performance. RESULTS AND CONCLUSION: To identify PGs with tumors from normal PGs, using combined T1WI and FS-T2WI, the deep learning model achieved an accuracy, sensitivity and specificity of 98.2% (497/506), 100% (119/119) and 97.7% (378/387), respectively, in the cross-validation set and 98.5% (67/68), 100% (20/20) and 97.9% (47/48), respectively, in the validation set. For patients with PGTs, automatic segmentation of PGTs on T1WI and FS-T2WI achieved mean dice coefficients of 86.1% and 84.2%, respectively, in the cross-validation set, and of 85.9% and 81.0%, respectively, in the validation set. The proposed deep learning model may assist the detection and segmentation of PGTs and, by acting as a second pair of eyes, ensure that incidentally detected PGTs on MRI are not missed.

Early detection of nasopharyngeal carcinoma: performance of a short contrast-free screening magnetic resonance imaging.

BACKGROUND: Although contrast-enhanced magnetic resonance imaging (MRI) detects early-stage nasopharyngeal carcinoma (NPC) not detected by endoscopic-guided biopsy (EGB), a short contrast-free screening MRI would be desirable for NPC screening programs. This study evaluated a screening MRI in a plasma Epstein-Barr virus (EBV)-DNA NPC screening program. METHODS: EBV-DNA-screen-positive patients underwent endoscopy, and endoscopy-positive patients underwent EGB. EGB was negative if the biopsy was negative or was not performed. Patients also underwent a screening MRI. Diagnostic performance was based on histologic confirmation of NPC in the initial study or during a follow-up period of at least 2 years. RESULTS: The study prospectively recruited 354 patients for MRI and endoscopy; 40/354 (11.3%) endoscopy-positive patients underwent EGB. Eighteen had NPC (5.1%), and 336 without NPC (94.9%) were followed up for a median of 44.8 months. MRI detected additional NPCs in 3/18 (16.7%) endoscopy-negative and 2/18 (11.1%) EGB-negative patients (stage I/II, n = 4; stage III, n = 1). None of the 24 EGB-negative patients who were MRI-negative had NPC. MRI missed NPC in 2/18 (11.1%), one of which was also endoscopy-negative. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of MRI, endoscopy, and EGB were 88.9%, 91.1%, 34.8%, 99.4%, and 91.0%; 77.8%, 92.3%, 35.0%, 98.7%, and 91.5%; and 66.7%, 92.3%, 31.6%, 98.1%, and 91.0%, respectively. CONCLUSION: A quick contrast-free screening MRI complements endoscopy in NPC screening programs. In EBV-screen-positive patients, MRI enables early detection of NPC that is endoscopically occult or negative on EGB and increases confidence that NPC has not been missed.

Dual-Level Augmentation Radiomics Analysis for Multisequence MRI Meningioma Grading.

BACKGROUND: Preoperative, noninvasive prediction of meningioma grade is important for therapeutic planning and decision making. In this study, we propose a dual-level augmentation strategy incorporating image-level augmentation (IA) and feature-level augmentation (FA) to tackle class imbalance and improve the predictive performance of radiomics for meningioma grading on Magnetic Resonance Imaging (MRI). METHODS: This study recruited 160 consecutive patients with pathologically proven meningioma (129 low-grade (WHO grade I) tumors; 31 high-grade (WHO grade II and III) tumors) with preoperative multisequence MRI imaging. A dual-level augmentation strategy combining IA and FA was applied and evaluated in 100 repetitions in 3-, 5-, and 10-fold cross-validation. RESULTS: The best area under the receiver operating characteristics curve of our method in 100 repetitions was ≥0.78 in all cross-validations. The corresponding cross-validation sensitivities (cross-validation specificity) were 0.72 (0.69), 0.76 (0.71), and 0.63 (0.82) in 3-, 5-, and 10-fold cross-validation, respectively. The proposed method achieved significantly better performance and distribution of results, outperforming single-level augmentation (IA or FA) or no augmentation in each cross-validation. CONCLUSIONS: The dual-level augmentation strategy using IA and FA significantly improves the performance of the radiomics model for meningioma grading on MRI, allowing better radiomics-based preoperative stratification and individualized treatment.

Radiomics Analysis in Characterization of Salivary Gland Tumors on MRI: A Systematic Review.

Radiomics analysis can potentially characterize salivary gland tumors (SGTs) on magnetic resonance imaging (MRI). The procedures for radiomics analysis were various, and no consistent performances were reported. This review evaluated the methodologies and performances of studies using radiomics analysis to characterize SGTs on MRI. We systematically reviewed studies published until July 2023, which employed radiomics analysis to characterize SGTs on MRI. In total, 14 of 98 studies were eligible. Each study examined 23-334 benign and 8-56 malignant SGTs. Least absolute shrinkage and selection operator (LASSO) was the most common feature selection method (in eight studies). Eleven studies confirmed the stability of selected features using cross-validation or bootstrap. Nine classifiers were used to build models that achieved area under the curves (AUCs) of 0.74 to 1.00 for characterizing benign and malignant SGTs and 0.80 to 0.96 for characterizing pleomorphic adenomas and Warthin's tumors. Performances were validated using cross-validation, internal, and external datasets in four, six, and two studies, respectively. No single feature consistently appeared in the final models across the studies. No standardized procedure was used for radiomics analysis in characterizing SGTs on MRIs, and various models were proposed. The need for a standard procedure for radiomics analysis is emphasized.

The Brain Tumor Segmentation (BraTS-METS) Challenge 2023: Brain Metastasis Segmentation on Pre-treatment MRI.

Clinical monitoring of metastatic disease to the brain can be a laborious and timeconsuming process, especially in cases involving multiple metastases when the assessment is performed manually. The Response Assessment in Neuro-Oncology Brain Metastases (RANO-BM) guideline, which utilizes the unidimensional longest diameter, is commonly used in clinical and research settings to evaluate response to therapy in patients with brain metastases. However, accurate volumetric assessment of the lesion and surrounding peri-lesional edema holds significant importance in clinical decision-making and can greatly enhance outcome prediction. The unique challenge in performing segmentations of brain metastases lies in their common occurrence as small lesions. Detection and segmentation of lesions that are smaller than 10 mm in size has not demonstrated high accuracy in prior publications. The brain metastases challenge sets itself apart from previously conducted MICCAI challenges on glioma segmentation due to the significant variability in lesion size. Unlike gliomas, which tend to be larger on presentation scans, brain metastases exhibit a wide range of sizes and tend to include small lesions. We hope that the BraTS-METS dataset and challenge will advance the field of automated brain metastasis detection and segmentation.

Repeatability of quantitative T1rho magnetic resonance imaging in normal brain tissues at 3.0T.

PURPOSE: T1rho imaging is a promising MRI technique for imaging of brain disease. This study aimed to assess the repeatability of quantitative T1rho imaging in the normal brain grey and white matter. METHODS: The study prospectively recruited 30 healthy volunteers without a history of neurological diseases or brain injury, and T1rho was performed and quantified from three imaging sessions. Repeat measures analysis of variance (ANOVA) and within-subject coefficients of variation (wCoV) was used to detect differences in T1rho values between the three scans. RESULTS: The results showed low wCoVs of less than 4.3% (range 0.92-4.27%) across all the brain structures. No significant differences were observed in T1rho measurement between the three scans (p > 0.05). The amygdala and hippocampus showed the highest T1rho values of 91.79 ± 2.55 msec and 91.07 ± 2.11 msec respectively, and the palladium and putamen had the lowest values of 67.60 ± 1.84 msec and 71.83 ± 1.85 msec respectively. CONCLUSION: T1rho shows high test-retest repeatability for whole brain imaging in serial imaging sessions, indicating it to be a reliable sequence for quantitative brain imaging.

Normal size of benign upper neck nodes on MRI: parotid, submandibular, occipital, facial, retroauricular and level IIb nodal groups.

PURPOSE: Nodal size is an important imaging criterion for differentiating benign from malignant nodes in the head and neck cancer staging. This study evaluated the size of normal nodes in less well-documented nodal groups in the upper head and neck on magnetic resonance imaging (MRI). METHODS: Analysis was performed on 289 upper head and neck MRIs of patients without head and neck cancer. The short axial diameters (SAD) of the largest node in the parotid, submandibular, occipital, facial, retroauricular and Level IIb of the upper internal jugular nodal groups were documented and compared to the commonly used threshold of ≥ 10 mm for diagnosis of a malignant node. RESULTS: Normal nodes in the parotid, occipital, retroauricular and Level IIb groups were small with a mean SAD ranging from 3.8 to 4.4 mm, nodes in the submandibular group were larger with a mean SAD of 5.5 mm and facial nodes were not identified. A size ≥ 10 mm was found in 0.8% of submandibular nodes. Less than 10% of the other nodal group had a SAD of ≥ 6 mm and none of them had a SAD ≥ 8 mm. CONCLUSION: To identify malignant neck nodes in these groups there is scope to reduce the size threshold of ≥ 10 mm to improve sensitivity without substantial loss of specificity.

Radiomics for Discriminating Benign and Malignant Salivary Gland Tumors; Which Radiomic Feature Categories and MRI Sequences Should Be Used?

The lack of a consistent MRI radiomic signature, partly due to the multitude of initial feature analyses, limits the widespread clinical application of radiomics for the discrimination of salivary gland tumors (SGTs). This study aimed to identify the optimal radiomics feature category and MRI sequence for characterizing SGTs, which could serve as a step towards obtaining a consensus on a radiomics signature. Preliminary radiomics models were built to discriminate malignant SGTs (n = 34) from benign SGTs (n = 57) on T1-weighted (T1WI), fat-suppressed (FS)-T2WI and contrast-enhanced (CE)-T1WI images using six feature categories. The discrimination performances of these preliminary models were evaluated using 5-fold-cross-validation with 100 repetitions and the area under the receiver operating characteristic curve (AUC). The differences between models' performances were identified using one-way ANOVA. Results show that the best feature categories were logarithm for T1WI and CE-T1WI and exponential for FS-T2WI, with AUCs of 0.828, 0.754 and 0.819, respectively. These AUCs were higher than the AUCs obtained using all feature categories combined, which were 0.750, 0.707 and 0.774, respectively (p < 0.001). The highest AUC (0.846) was obtained using a combination of T1WI + logarithm and FS-T2WI + exponential features, which reduced the initial features by 94.0% (from 1015 × 3 to 91 × 2). CE-T1WI did not improve performance. Using one feature category rather than all feature categories combined reduced the number of initial features without compromising radiomic performance.

Editorial Comment: Iron-sensitive MR imaging of the primary motor cortex to differentiate hereditary spastic paraplegia from other motor neuron diseases.

KEY POINTS: • Conventional and advanced MR techniques may aid in the diagnosis of motor neuron disease.• Iron-sensitive MR imaging of the primary motor cortex may reveal changes to help differentiate hereditary spastic paraplegia (HSP) from UMM predominant amyotrophic lateral sclerosis (UMN-ALS) and primary lateral sclerosis (PLS).• Additional research in this area is necessary.

Lower Prevalence and Severity of Degenerative Changes in the Lumbar Spine in Elderly Hong Kong Chinese Compared With Age-Matched Italian Caucasian Women.

STUDY DESIGN: Cross-sectional observational study. OBJECTIVE: The aim was to compare the prevalence and severity of radiographic lumbar spine degeneration between elderly Hong Kong Chinese and elderly Italian Caucasian women. SUMMARY OF BACKGROUND DATA: Rates of symptomatic low back pain and osteoporotic vertebral fracture have been shown to be lower in Asian and Chinese populations compared with Caucasians, but ethnic differences in spinal degeneration are less established. METHODS: Lumbar spine lateral radiographs of 566 age-matched (mean: 73.6 yr; range: 65-87 yr) female subjects from two population-based epidemiological studies from Hong Kong (n=283) and Rome, Italy (n=283) were reviewed. Grading of degeneration categories: disk height loss (none, <30%, 30%-60%, >60%), osteophyte formation (not present, minimal, small, large), endplate sclerosis (none, mild, moderate, severe), and antero/retrolisthesis (none, <25%, 25%-50%, >50%) was performed for vertebral levels from L1/2 to L5/S1 (five levels). Each category was assigned a score (0, 1, 2, 3) at individual vertebral level according to severity. The total degeneration score was obtained by adding scores for all categories across the vertebral levels. RESULTS: Italian subjects [total score (mean±SD): 7.0±5.5] had a higher severity of overall degenerative changes compared with Hong Kong subjects (5.7±4.4), P <0.01. Italian subjects had higher scores for individual findings of disk height loss (Italian, 3.6±2.8 vs. Hong Kong 2.5±2.1, P <0.01); antero/retrolisthesis (Italian 0.3±0.7 vs. Hong Kong 0.2±0.4, P =0.01); and endplate sclerosis (Italian 1.0±1.2 vs. Hong Kong 0.6±1.0, P <0.01). At each individual level from L1/2 to L5/S1, total degeneration scores were higher in Italian than Hong Kong subjects ( P <0.01-0.04). CONCLUSION: Degenerative changes in the lumbar spine are less prevalent and less severe in elderly Hong Kong Chinese women than in age-matched Italian Caucasian women. The observed differences may reflect a foundational background influence of genetic predisposition that requires further studies.

Prognostic value of cervical nodal necrosis on staging imaging of nasopharyngeal carcinoma in era of intensity-modulated radiotherapy: a systematic review and meta-analysis.

PURPOSES: To systematically review and perform meta-analysis to evaluate the prognostic value of cervical nodal necrosis (CNN) on the staging computed tomography/magnetic resonance imaging (MRI) of nasopharyngeal carcinoma (NPC) in era of intensity-modulated radiotherapy. METHODS: Literature search through PubMed, EMBASE, and Cochrane Library was conducted. The hazard ratios (HRs) with 95% confidence intervals (CIs) of CNN for distant metastasis-free survival (DMFS), disease free survival (DFS) and overall survival (OS) were extracted from the eligible studies and meta-analysis was performed to evaluate the pooled HRs with 95%CI. RESULTS: Nine studies, which investigated the prognostic values of 6 CNN patterns on MRI were included. Six/9 studies were eligible for meta-analysis, which investigated the CNN presence/absence in any nodal group among 4359 patients. The pooled unadjusted HRs showed that the CNN presence predicted poor DMFS (HR =1.89, 95%CI =1.72-2.08), DFS (HR =1.57, 95%CI =1.08-2.26), and OS (HR =1.87, 95%CI =1.69-2.06). The pooled adjusted HRs also showed the consistent results for DMFS (HR =1.34, 95%CI =1.17-1.54), DFS (HR =1.30, 95%CI =1.08-1.56), and OS (HR =1.61, 95%CI =1.27-2.04). Results shown in the other studies analysing different CNN patterns indicated the high grade of CNN predicted poor outcome, but meta-analysis was unable to perform because of the heterogeneity of the analysed CNN patterns. CONCLUSION: The CNN observed on the staging MRI is a negative factor for NPC outcome, suggesting that the inclusion of CNN is important in the future survival analysis. However, whether and how should CNN be included in the staging system warrant further evaluation.

Serial changes in liquid biopsy-derived variant allele frequency predict immune checkpoint inhibitor responsiveness in the pan-cancer setting.

Major immunotherapy challenges include a limited number of predictive biomarkers and the unusual imaging features post-therapy, such as pseudo-progression, which denote immune infiltrate-mediated tumor enlargement. Such phenomena confound clinical decision-making, since the cancer may eventually regress, and the patient should stay on treatment. We prospectively evaluated serial, blood-derived cell-free DNA (cfDNA) (baseline and 2-3 weeks post-immune checkpoint inhibitors [ICIs]) for variant allele frequency (VAF) and blood tumor mutation burden (bTMB) changes (next-generation sequencing) (N = 84 evaluable patients, diverse cancers). Low vs. high cfDNA-derived average adjusted ΔVAF (calculated by a machine-learning model) was an independent predictor of higher clinical benefit rate (stable disease ≥6 months/complete/partial response) (69.2% vs. 22.5%), and longer median progression-free (10.1 vs. 2.25 months) and overall survival (not reached vs. 6.1 months) (all P < .001, multivariate). bTMB changes did not correlate with outcomes. Therefore, early dynamic changes in cfDNA-derived VAF were a powerful predictor of pan-cancer immunotherapy outcomes.

Current Concepts in Imaging Diagnosis and Screening of Blunt Cerebrovascular Injuries.

Blunt cerebrovascular injury (BCVI) is an often underrecognized injury occurring in the carotid or vertebral arteries, associated with a risk of ischemic stroke and potential for poor neurological outcome or death. Computed tomographic angiography (CTA) is the most common modality for initial screening and diagnosis. Vessel wall intimal injuries, intraluminal thrombus, dissection, intramural hematoma, pseudoaneurysm, vessel transection, and arteriovenous fistula, are potential findings to be considered in approach to imaging. Identification of high-risk trauma patients based on clinical and radiological risk factors can determine patients at risk of BCVI for targeted screening.

Machine learning application for the prediction of SARS-CoV-2 infection using blood tests and chest radiograph.

Triaging and prioritising patients for RT-PCR test had been essential in the management of COVID-19 in resource-scarce countries. In this study, we applied machine learning (ML) to the task of detection of SARS-CoV-2 infection using basic laboratory markers. We performed the statistical analysis and trained an ML model on a retrospective cohort of 5148 patients from 24 hospitals in Hong Kong to classify COVID-19 and other aetiology of pneumonia. We validated the model on three temporal validation sets from different waves of infection in Hong Kong. For predicting SARS-CoV-2 infection, the ML model achieved high AUCs and specificity but low sensitivity in all three validation sets (AUC: 89.9-95.8%; Sensitivity: 55.5-77.8%; Specificity: 91.5-98.3%). When used in adjunction with radiologist interpretations of chest radiographs, the sensitivity was over 90% while keeping moderate specificity. Our study showed that machine learning model based on readily available laboratory markers could achieve high accuracy in predicting SARS-CoV-2 infection.

Federated deep learning for detecting COVID-19 lung abnormalities in CT: a privacy-preserving multinational validation study.

Data privacy mechanisms are essential for rapidly scaling medical training databases to capture the heterogeneity of patient data distributions toward robust and generalizable machine learning systems. In the current COVID-19 pandemic, a major focus of artificial intelligence (AI) is interpreting chest CT, which can be readily used in the assessment and management of the disease. This paper demonstrates the feasibility of a federated learning method for detecting COVID-19 related CT abnormalities with external validation on patients from a multinational study. We recruited 132 patients from seven multinational different centers, with three internal hospitals from Hong Kong for training and testing, and four external, independent datasets from Mainland China and Germany, for validating model generalizability. We also conducted case studies on longitudinal scans for automated estimation of lesion burden for hospitalized COVID-19 patients. We explore the federated learning algorithms to develop a privacy-preserving AI model for COVID-19 medical image diagnosis with good generalization capability on unseen multinational datasets. Federated learning could provide an effective mechanism during pandemics to rapidly develop clinically useful AI across institutions and countries overcoming the burden of central aggregation of large amounts of sensitive data.

Test-retest repeatability of T1rho (T1ρ) MR imaging in the head and neck.

PURPOSE: T1rho imaging is a new quantitative MRI sequence for head and neck cancer and the repeatability for this region is unknown. This study aimed to evaluate the repeatability of quantitative T1rho imaging in the head and neck. MATERIALS AND METHODS: T1rho imaging of the head and neck was prospectively performed in 15 healthy participants on three occasions. Scan 1 and 2 were performed with a time interval of 30 minutes (intra-session) and scan 3 was performed 14 days later (inter-session). T1rho values for normal tissues (parotid glands, palatine tonsils, pterygoid muscles, and tongue) were obtained on each scan. Intra-class coefficients (ICCs), within-subject coefficient of variances (wCoVs), and repeatability coefficient (RCs) of the intra-session scan (scan 1 vs 2) and inter-session scan (scan 1 vs 3) for the normal tissues were calculated. RESULTS: The ICCs of T1rho values for normal tissues were almost perfect (0.83-0.97) for intra-session scans and were substantial (0.71-0.80) for inter-session scans. The wCoVs showed a small range (2.46%-3.30%) for intra-session scans, and slightly greater range (3.27%-6.51%) for inter-session scan. The greatest and lowest wCoVs of T1rho were found in the parotid gland and muscles, respectively. The T1rho RCs varied for all tissues between intra- and inter- sessions, and the greatest RC of 10.07 msec was observed for parotid gland on inter-session scan. CONCLUSION: T1rho imaging is a repeatable quantitative MRI sequence in the head and neck but variances of T1rho values among tissues should be take into account during analysis.

Intravoxel incoherent motion diffusion-weighted imaging for discrimination of benign and malignant retropharyngeal nodes.

PURPOSE: Anatomical imaging criteria for the diagnosis of malignant head and neck nodes may not always be reliable. This study aimed to evaluate the diagnostic value of conventional diffusion-weighted imaging (DWI) and intravoxel incoherent motion (IVIM) DWI in discriminating benign and malignant metastatic retropharyngeal nodes (RPNs). METHODS: IVIM DWI using 14 b-values was performed on RPNs of 30 patients with newly diagnosed metastatic nasopharyngeal carcinoma (NPC) and 30 patients with elevated plasma Epstein-Barr virus (EBV)-DNA without NPC who were part of an EBV-based NPC screening program. Histogram measurements of the two groups were compared for pure diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion volume fraction (f) and apparent diffusion coefficient (ADC) using the Mann-Whitney U test. Area under the curves (AUCs) of significant measurements were calculated from receiver-operating characteristics analysis and compared using the DeLong test. RESULTS: Compared with metastatic RPNs, benign RPNs had lower ADCmean (0.73 vs 0.82 × 10-3 mm2/s) and Dmean (0.60 vs 0.71 × 10-3 mm2/s) and a higher D*mean (35.21 vs 28.66 × 10-3 mm2/s) (all p < 0.05). There was no difference in the f measurements between the two groups (p = 0.204 to 0.301). Dmean achieved the highest AUC of 0.800, but this was not statistically better than the AUCs of the other parameters (p = 0.148 to 0.991). CONCLUSION: Benign RPNs in patients with EBV-DNA showed greater restriction of diffusion compared with malignant metastatic RPNs from NPC. IVIM did not show a significant advantage over conventional DWI in discriminating benign and malignant nodes.

Pre-treatment intravoxel incoherent motion diffusion-weighted imaging predicts treatment outcome in nasopharyngeal carcinoma.

PURPOSE: To evaluate whether pre-treatment intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) can predict treatment outcome after 2 years in patients with nasopharyngeal carcinoma (NPC). METHOD: One hundred and sixty-one patients with newly diagnosed NPC underwent pre-treatment IVIM-DWI. Univariate Cox regression analysis was performed to evaluate the correlation of the mean values of the pure diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion fraction and apparent diffusion coefficient with local relapse-free survival (LRFS), regional relapse-free survival (RRFS), distant metastases-free survival (DMFS) and disease-free survival (DFS). Significant diffusion parameters, together with staging, age, gender and treatment as confounding factors, were added into a multivariate model. The area under the curves (AUCs) of significant parameters for disease relapse were compared using the Delong test. RESULTS: Disease relapse occurred in 30 % of the patients at a median follow-up time of 52.1 months. The multivariate analysis showed that high D and T-staging were correlated with poor LRFS (p = 0.042 and 0.020, respectively) and poor DFS (p = 0.023 and 0.001, respectively); low D* and high T-staging with poor RRFS (p = 0.020 and 0.033, respectively); and high N-staging with poor DMFS (p = 0.006). D with the optimal threshold of ≥0.68 × 10-3 mm2/s and T-staging showed similar AUCs (AUC = 0.614 and 0.651, respectively; p = 0.493) for predicting disease relapse. CONCLUSION: High D and low D* were predictors of poor locoregional outcome but none of the diffusion parameters predicted DMFS in NPC.