Contrast-enhanced T1-weighted MR Research Articles

Morphological MRI features as prognostic indicators in brain metastases

BackgroundStereotactic radiotherapy is the preferred treatment for managing patients with fewer than five brain metastases (BMs). However, some lesions recur after irradiation. The purpose of this study was to identify patients who are at a higher risk of failure, which can help in adjusting treatments and preventing recurrence.MethodsIn this retrospective multicenter study, we analyzed the predictive significance of a set of interpretable morphological features derived from contrast-enhanced (CE) T1-weighted MR images as imaging biomarkers using Kaplan–Meier analysis. The feature sets studied included the total and necrotic volumes, the surface regularity and the CE rim width. Additionally, we evaluated other nonmorphological variables and performed multivariate Cox analysis.ResultsA total of 183 lesions in 128 patients were included (median age 61 [31–95], 64 men and 64 women) treated with stereotactic radiotherapy (57% single fraction, 43% fractionated radiotherapy). None of the studied variables measured at diagnosis were found to have prognostic value. However, the total and necrotic volumes and the CE rim width measured at the first follow-up after treatment and the change in volume due to irradiation can be used as imaging biomarkers for recurrence. The optimal classification was achieved by combining the changes in tumor volume before and after treatment with the presence or absence of necrosis (p < < 0.001).ConclusionThis study demonstrated the prognostic significance of interpretable morphological features extracted from routine clinical MR images following irradiation in brain metastases, offering valuable insights for personalized treatment strategies.

Multiparameter MRI-based radiomics nomogram for preoperative prediction of brain invasion in atypical meningioma:a multicentre study

ObjectiveTo develop a nomogram based on tumor and peritumoral edema (PE) radiomics features extracted from preoperative multiparameter MRI for predicting brain invasion (BI) in atypical meningioma (AM).MethodsIn this retrospective study, according to the 2021 WHO classification criteria, a total of 469 patients with pathologically confirmed AM from three medical centres were enrolled and divided into training (n = 273), internal validation (n = 117) and external validation (n = 79) cohorts. BI was diagnosed based on the histopathological examination. Preoperative contrast-enhanced T1-weighted MR images (T1C) and T2-weighted MR images (T2) for extracting meningioma features and T2-fluid attenuated inversion recovery (FLAIR) sequences for extracting meningioma and PE features were obtained. The multiple logistic regression was applied to develop separate multiparameter radiomics models for comparison. A nomogram was developed by combining radiomics features and clinical risk factors, and the clinical usefulness of the nomogram was verified using decision curve analysis.ResultsAmong the clinical factors, PE volume and PE/tumor volume ratio are the risk of BI in AM. The combined nomogram based on multiparameter MRI radiomics features of meningioma and PE and clinical indicators achieved the best performance in predicting BI in AM, with area under the curve values of 0.862 (95% CI, 0.819–0.905) in the training cohort, 0.834 (95% CI, 0.780–0.908) in the internal validation cohort and 0.867 (95% CI, 0.785–0.950) in the external validation cohort, respectively.ConclusionsThe nomogram based on tumor and PE radiomics features extracted from preoperative multiparameter MRI and clinical factors can predict the risk of BI in patients with AM.

Synthesizing Contrast-Enhanced MR Images from Noncontrast MR Images Using Deep Learning.

Recent developments in deep learning methods offer a potential solution to the need for alternative imaging methods due to concerns about the toxicity of gadolinium-based contrast agents. The purpose of the study was to synthesize virtual gadolinium contrast-enhanced T1-weighted MR images from noncontrast multiparametric MR images in patients with primary brain tumors by using deep learning. We trained and validated a deep learning network by using MR images from 335 subjects in the Brain Tumor Segmentation Challenge 2019 training data set. A held out set of 125 subjects from the Brain Tumor Segmentation Challenge 2019 validation data set was used to test the generalization of the model. A residual inception DenseNet network, called T1c-ET, was developed and trained to simultaneously synthesize virtual contrast-enhanced T1-weighted (vT1c) images and segment the enhancing portions of the tumor. Three expert neuroradiologists independently scored the synthesized vT1c images by using a 3-point Likert scale, evaluating image quality and contrast enhancement against ground truth T1c images (1 = poor, 2 = good, 3 = excellent). The synthesized vT1c images achieved structural similarity index, peak signal-to-noise ratio, and normalized mean square error scores of 0.91, 64.35, and 0.03, respectively. There was moderate interobserver agreement between the 3 raters, regarding the algorithm's performance in predicting contrast enhancement, with a Fleiss kappa value of 0.61. Our model was able to accurately predict contrast enhancement in 88.8% of the cases (scores of 2 to 3 on the 3-point scale). We developed a novel deep learning architecture to synthesize virtual postcontrast enhancement by using only conventional noncontrast brain MR images. Our results demonstrate the potential of deep learning methods to reduce the need for gadolinium contrast in the evaluation of primary brain tumors.

Development and validation of a machine learning algorithm for predicting diffuse midline glioma, H3 K27-altered, H3 K27 wild-type high-grade glioma, and primary CNS lymphoma of the brain midline in adults.

Preoperative diagnosis of diffuse midline glioma, H3 K27-altered (DMG-A) and midline high-grade glioma without H3 K27 alteration (DMG-W), as well as midline primary CNS lymphoma (PCNSL) in adults, is challenging but crucial. The aim of this study was to develop a model for predicting these three entities using machine learning (ML) algorithms. Thirty-three patients with DMG-A, 35 with DMG-W, and 35 with midline PCNSL were retrospectively enrolled in the study. Radiomics features were extracted from contrast-enhanced T1-weighted MR images. Two radiologists evaluated the conventional MRI features of the tumors, such as shape. Patient age, tumor volume, and conventional MRI features were considered clinical features. The data set was randomly stratified into 70% training and 30% testing cohorts. Predictive models based on the clinical features, radiomics features, and integration of clinical and radiomics features were established through ML. The performances of the models were evaluated by calculating the area under the receiver operating characteristic curve, accuracy, sensitivity, and specificity. Subsequently, 10 patients with DMG-A, 10 with DMG-W, and 12 with PCNSL were enrolled from another institution to validate the established models. The predictive models based on clinical features, radiomics features, and the integration of clinical and radiomics features through the support vector machine algorithm had the optimal accuracies in the training, testing, and validation cohorts, and the accuracies in the testing cohort were 0.871, 0.892, and 0.903, respectively. Age, 2 radiomics features, and 3 conventional MRI features were the 6 most significant features in the established integrated model. The integrated prediction model established by ML provides high discriminatory accuracy for predicting DMG-A, DMG-W, and midline PCNSL in adults.

Evolution of non-perfused volume after transurethral ultrasound ablation of prostate: A retrospective 12-month analysis

BackgroundA detailed understanding of the non-perfused volume (NPV) evolution after prostate ablation therapy is lacking. The impact of different diseased prostate tissues on NPV evolution post-ablation is unknown. PurposeTo characterize the NPV evolution for three treatment groups undergoing heat-based prostate ablation therapy, including benign prostatic hyperplasia (BPH), primary prostate cancer (PCa), and radiorecurrent PCa. Materials and methodsStudy design and data analysis were performed retrospectively. All patients received MRI-guided transurethral ultrasound ablation (TULSA). 21 BPH, 28 radiorecurrent PCa and 40 primary PCa patients were included. Using the T1-weighted contrast-enhanced MR image, the NPV was manually contoured by an experienced radiologist. All patients received an MRI immediately following the ablation. Follow-up included MRI at 3- and 12 months for BPH and radiorecurrent PCa patients and at 6- and 12 months for primary PCa patients. ResultsA significant difference between BPH and radiorecurrent PCa patients was observed at three months (p < 0.0001, Wilcoxon rank sum test), with the median NPV decreasing by 77 % for BPH patients but increasing by 4 % for radiorecurrent PCa patients. At six months, the median NPV decreased by 97 % for primary PCa. Across all groups, although 40 % of patients had residual NPV at 12 months, it tended to be < 1 mL. ConclusionThe resolution of necrotic tissue after ablation was markedly slower for irradiated than treatment-naïve prostate tissue. These results may account for the increased toxicity observed after radiorecurrent salvage therapy. By 12 months, most necrotic prostate tissue had disappeared in every treatment group.

Explainable machine learning via intra-tumoral radiomics feature mapping for patient stratification in adjuvant chemotherapy for locoregionally advanced nasopharyngeal carcinoma.

This study aimed to discover intra-tumor heterogeneity signature and validate its predictive value for adjuvant chemotherapy (ACT) following concurrent chemoradiotherapy (CCRT) in locoregionally advanced nasopharyngeal carcinoma (LA-NPC). 397 LA-NPC patients were retrospectively enrolled. Pre-treatment contrast-enhanced T1-weighted (CET1-w) MR images, clinical variables, and follow-up were retrospectively collected. We identified single predictive radiomic feature from primary gross tumor volume (GTVnp) and defined predicted subvolume by calculating voxel-wised feature mapping and within GTVnp. We independently validate predictive value of identified feature and associated predicted subvolume. Only one radiomic feature, gldm_DependenceVariance in 3mm-sigma LoG-filtered image, was discovered as a signature. In the high-risk group determined by the signature, patients received CCRT + ACT achieved 3-year disease free survival (DFS) rate of 90% versus 57% (HR, 0.20; 95%CI, 0.05-0.94; P = 0.007) for CCRT alone. The multivariate analysis showed patients receiving CCRT + ACT had a HR of 0.21 (95%CI: 0.06-0.68, P = 0.009) for DFS compared to those receiving CCRT alone. The predictive value can also be generalized to the subvolume with multivariate HR of 0.27 (P = 0.017) for DFS. The signature with its heterogeneity mapping could be a reliable and explainable ACT decision-making tool in clinical practice.

Uterine Inversion.

HomeRadioGraphicsVol. 43, No. 6 PreviousNext Cases from the Cooky JarFree AccessUterine InversionKaitlin M. Zaki-Metias , Melina Hosseiny, Fardad Behzadi, Patricia BalthazarKaitlin M. Zaki-Metias , Melina Hosseiny, Fardad Behzadi, Patricia BalthazarAuthor AffiliationsFrom the Department of Radiology, Trinity Health Oakland Hospital/Wayne State University School of Medicine, 44405 Woodward Ave, Medical Education, Pontiac, MI 48341 (K.M.Z.M.); Department of Radiology, University of California San Diego, San Diego, Calif (M.H.); Department of Radiology, Brigham and Women's Hospital/Harvard Medical School, Boston, Mass (F.B.); and Department of Radiology and Imaging Sciences, Emory University School of Medicine, Atlanta, Ga (P.B.).Address correspondence to K.M.Z.M. (email: [email protected]).Kaitlin M. Zaki-Metias Melina HosseinyFardad BehzadiPatricia BalthazarPublished Online:May 18 2023https://doi.org/10.1148/rg.230004MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In Uterine inversion, or uterine intussusception, is a rare, potentially life-threatening entity characterized by inversion of the fundus into the uterine cavity, which often manifests as an acute complication of childbirth (1,2). Less commonly, uterine inversion is nonpuerperal, occurring as a result of an endometrial or myometrial mass (2,3). Patients typically present with symptoms that include vaginal bleeding and pelvic pain.At MRI, uterine inversion appears as a U-shaped uterus with loss of the normal convex fundal contour on sagittal images (Fig 1A, 1B) (2). A target configuration may be visualized on axial images, representing the concentric serosa, myometrium, and endometrium (Fig 1C, 1D) (2). MRI is also useful in assessing the cause in patients with nonpuerperal uterine inversion, including submucosal leiomyomas, uterine leiomyosarcomas, and endometrial malignancies. There are four grades of uterine inversion, which are characterized by the degree of inferior bowing of the fundus relative to the cervix and/or vaginal introitus (Fig 2).Figure 1. Uterine inversion in a previously healthy premenopausal female patient with acute onset of pelvic pain and vaginal bleeding. A bleeding vaginal or cervical mass was suspected on physical examination, and pelvic MRI was performed for further evaluation. Sagittal T2-weighted (A) and contrast-enhanced fat-suppressed T1-weighted (B) MR images of the pelvis demonstrate a U-shaped uterus (U) with loss of the normal fundal contour, with protrusion of the fundus into the distended vaginal canal (V), consistent with grade III uterine inversion. The cervix (C) is located superior to the inverted fundus. At the fundus, there is an ill-defined rounded lesion (arrow in A and B) with heterogeneous T1 signal intensity and enhancement with irregular margins, with adjacent heterogeneous T2 signal intensity and increased T1 signal intensity. The findings are most consistent with a hemorrhagic submucosal uterine leiomyoma. Axial T2-weighted (C) and contrast-enhanced fat-suppressed T1-weighted (D) MR images of the pelvis show concentric rings of signal intensity indicating the serosa, myometrium, endometrium, and vagina.Figure 1.Download as PowerPointOpen in Image Viewer Figure 2. Illustration depicts the normal anatomic position of the uterus (A) and the various grades of inversion. Grade I uterine inversion (B) is characterized by inferior bowing of the fundus up to but not beyond the cervix. In grade II uterine inversion (C), the inverted fundus protrudes through the cervix and into the vagina. The uterus is fully inverted and protrudes out of the vagina in grade III uterine inversion (D).Figure 2.Download as PowerPointOpen in Image Viewer Disclosures of conflicts of interest.—K.M.Z.M. Meeting attendance and travel reimbursement from Trinity Health Oakland Hospital. P.B. Editorial board member of RadioGraphics, Association of University Radiologists GE Radiology Research Academic Fellowship, meeting travel reimbursement from the American College of Radiology.K.M.Z.M and P.B. have provided disclosures (see end of article); all other authors have disclosed no relevant relationships.

Single Cerebral Metastasis Mimicking Pyogenic Abscess in a Patient with Lung Adenocarcinoma.

HomeRadiology: Imaging CancerVol. 5, No. 3 Images in CancerFree AccessSingle Cerebral Metastasis Mimicking Pyogenic Abscess in a Patient with Lung AdenocarcinomaVíctor Pérez-Riverola , Olalla Vázquez Muiños, Mario Prenafeta Moreno, Roser Monmany Badia, Khalid El-Hamshari Rebollo, Esther Granell MorenoVíctor Pérez-Riverola , Olalla Vázquez Muiños, Mario Prenafeta Moreno, Roser Monmany Badia, Khalid El-Hamshari Rebollo, Esther Granell MorenoAuthor AffiliationsFrom the Department of Neuroradiology, UDIAT-CD, Hospital Universitari Parc Taulí, Parc del Taulí 1, 08208 Sabadell, Spain.Address correspondence to V.P.R. (email: [email protected]).Víctor Pérez-Riverola Olalla Vázquez MuiñosMario Prenafeta MorenoRoser Monmany BadiaKhalid El-Hamshari RebolloEsther Granell MorenoPublished Online:Apr 7 2023https://doi.org/10.1148/rycan.230004MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In As clinical signs of infection are often absent in patients with pyogenic abscesses, differential diagnosis with cerebral metastases is based on MRI findings.However, both entities often manifest as ring-enhancing lesions. Pyogenic brain abscesses usually show restricted diffusion in the lesion core and the dual-rim sign at susceptibility-weighted imaging. The dual-rim sign consists of two concentric rims bordering the lesion: a complete, smooth outer hypointense rim representing the fibrocollagenous capsular wall and an inner hyperintense rim representing a zone of granulation tissue (1). At perfusion imaging, the capsule of a cerebral abscess has a lower relative cerebral blood volume than the necrotic or cystic peripheral wall of a tumor due to microvessel angiogenesis by the tumor (2).Some cerebral metastases also show restricted diffusion, which does not correlate with the histologic type of the primary cancer (3). Proposed mechanisms for restricted diffusion in metastases include intracellular edema in viable tumor cells during transition to complete liquefactive necrosis (4) and a highly viscous mucin concentration (5).Key findings that help distinguish these atypical metastases from pyogenic abscess include high peripheral wall relative cerebral blood volume and absence of the dual-rim sign (Figure).Brain MR images in a 37-year-old male patient, a heavy tobacco and cannabis smoker, after a first seizure. (A) Contrast-enhanced T1-weighted MR image shows a smooth, ring-enhancing nodular right frontal lobe lesion surrounded by abundant vasogenic edema. (B, C) On diffusion-weighted images, the central core shows restricted diffusion. (D) On susceptibility-weighted image, a hypointense smooth ring completely surrounds the lesion, simulating a capsule. (E) The relative cerebral blood volume of the lesion’s peripheral wall (arrows) is high (mean relative cerebral blood volume ratio: 7.61 ± 2.70 [SD]). Chest CT findings (not shown) detected a cavitary pulmonary mass. Tests for tuberculosis and bronchoalveolar lavage fluid analysis were negative. Chest CT findings were initially misinterpreted as an infectious process. Follow-up brain MRI 8 weeks after starting antibiotics showed enlargement of the frontal lobe lesion, new ring-enhancing cerebral and cerebellar lesions, and signs of leptomeningeal carcinomatosis in the posterior fossa. The patient was diagnosed with lung adenocarcinoma following lung lesion biopsy.Download as PowerPointOpen in Image Viewer Disclosures of conflicts of interest: V.P.R. No relevant relationships. O.V.M. No relevant relationships. M.P.M. No relevant relationships. R.M.B. No relevant relationships. K.E.H.R. No relevant relationships. E.G.M. No relevant relationships.Keywords: MRI, MR-Diffusion-weighted Imaging, MR-Perfusion, Metastasis, Neuro-oncology, Lung Adenocarcinoma, Pyogenic AbscessAuthors declared no funding for this work.

Radiomic feature repeatability and its impact on prognostic model generalizability: A multi-institutional study on nasopharyngeal carcinoma patients

Background and purposeTo investigate the radiomic feature (RF) repeatability via perturbation and its impact on cross-institutional prognostic model generalizability in Nasopharyngeal Carcinoma (NPC) patients. Materials and methods286 and 183 NPC patients from two institutions were included for model training and validation. Perturbations with random translations and rotations were applied to contrast-enhanced T1-weighted (CET1-w) MR images. RFs were extracted from primary tumor volume under a wide range of image filtering and discretization settings. RF repeatability was assessed by intraclass correlation coefficient (ICC), which was used to equally separate the RFs into low- and high-repeatable groups by the median value. After feature selection, multivariate Cox regression and Kaplan-Meier analysis were independently employed to develop and analyze prognostic models. Concordance index (C-index) and P-value from log-rank test were used to assess model performance. ResultsMost textural RFs from high-pass wavelet-filtered images were susceptible to image perturbations. It was more prominent when a smaller discretization bin number was used (e.g., 8, mean ICC = 0.69). Using high-repeatable RFs for model development yielded a significantly higher C-index (0.63) in the validation cohort than when only low-repeatable RFs were used (0.57, P = 0.024), suggesting higher model generalizability. Besides, significant risk stratification in the validation cohort was observed only when high-repeatable RFs were used (P < 0.001). ConclusionRepeatability of RFs from high-pass wavelet-filtered CET1-w MR images of primary NPC tumor was poor, particularly when a smaller bin number was used. Exclusive use of high-repeatable RFs is suggested to safeguard model generalizability for wide-spreading clinical utilization.

Prognostic value of PET/CT and MR-based baseline radiomics among patients with non-metastatic nasopharyngeal carcinoma

Radiomics is an emerging imaging assessment technique that has shown promise in predicting survival among nasopharyngeal carcinoma (NPC) patients. Studies so far have focused on PET or MR-based radiomics independently. The aim of our study was to evaluate the prognostic value of clinical and radiomic parameters derived from both PET/CT and MR. Retrospective evaluation of 124 NPC patients with PET/CT and radiotherapy planning MR (RP-MR). Primary tumors were segmented using dedicated software (LIFEx version 6.1) from PET, CT, contrast-enhanced T1-weighted (T1-w), and T2-weighted (T2-w) MR sequences with 376 radiomic features extracted. Summary statistics describe patient, disease, and treatment characteristics. The Kaplan-Meier (KM) method estimates overall survival (OS) and progression-free survival (PFS). Clinical factors selected based on univariable analysis and the multivariable Cox model were subsequently constructed with radiomic features added. The final models comparing clinical, clinical + RP-MR, clinical + PET/CT and clinical + RP-MR + PET/CT for OS and PFS demonstrated that combined radiomic signatures were significantly associated with improved survival prognostication (AUC 0.62 vs 0.81 vs 0.75 vs 0.86 at 21 months for PFS and 0.56 vs 0.85 vs 0.79 vs 0.96 at 24 months for OS). Clinical + RP-MR features initially outperform clinical + PET/CT for both OS and PFS (<18 months), and later in the clinical course for PFS (>42 months). Our study demonstrated that PET/CT-based radiomic features may improve survival prognostication among NPC patients when combined with baseline clinical and MR-based radiomic features.

NEIM-06 COMBINING CLINICAL VARIABLES AND RADIOMIC FEATURES TO HELP DISTINGUISH RADIATION NECROSIS FROM TUMOR IN PATIENTS WITH MELANOMA BRAIN METASTASES TREATED WITH RADIOSURGERY

Abstract BACKGROUND Following Gamma Knife SRS (GK-SRS), the conventional imaging characteristics of radiation necrosis (RN) mimic those of tumor progression, introducing considerable uncertainty in diagnosis. Previous studies have identified clinical variables associated with RN; however, diagnosis primarily relied on interpretation of imaging with only a minority confirmed using the gold standard of pathological examination. Furthermore, the cohorts of these studies included a mix of primary histologies. PURPOSE To identify the combination of clinical variables and radiomic features most predictive of RN in patients with melanoma brain metastasis (BM) with GK-SRS in order to train a machine learning classifier to distinguish RN from tumor progression. METHODS We retrospectively studied 86 patients with a melanoma BM that received initial GK-SRS followed by resection, thereby pathologically confirming tumor or RN. Clinical variables including lesion volume, age at surgery, GK-SRS dose, lesion hemorrhage, lesion location, gender, BM velocity, and drug therapy type were obtained from chart review. We extracted radiomic features from contrast-enhanced T1-weighted MR images using PyRadiomics. A consensus clustering algorithm identified representative radiomic features. Non-parametric hypothesis testing was performed on the clinical variables and representative radiomic features. RESULTS Of the 86 patients, 17 (19.8%) patients exhibited RN and 69 exhibited tumor progression. Lesion volume was associated with development of RN (p = 0.038&amp;lt;0.05) with a median volume of 1.5 cc (0.01-26.71 cc). Clustering analysis identified seven representative radiomic features; five were found to have statistically significant association with development of RN. CONCLUSION In this dataset with pathologically confirmed diagnoses in a histologically homogeneous patient cohort, we reproduced previously reported findings that the clinical variable of lesion volume is associated with RN and we identified several radiomic features associated with RN in patients with melanoma BM. We are using these variables and features to train a machine learning classifier to distinguish RN from tumor.

2D probabilistic undersampling pattern optimization for MR image reconstruction.

With 3D magnetic resonance imaging (MRI), a tradeoff exists between higher image quality and shorter scan time. One way to solve this problem is to reconstruct high-quality MRI images from undersampled k-space. There have been many recent studies exploring effective k-space undersampling patterns and designing MRI reconstruction methods from undersampled k-space, which are two necessary steps. Most studies separately considered these two steps, although in theory, their performance is dependent on each other. In this study, we propose a joint optimization model, trained end-to-end, to simultaneously optimize the undersampling pattern in the Fourier domain and the reconstruction model in the image domain. A 2D probabilistic undersampling layer was designed to optimize the undersampling pattern and probability distribution in a differentiable manner. A 2D inverse Fourier transform layer was implemented to connect the Fourier domain and the image domain during the forward and back propagation. Finally, we discovered an optimized relationship between the probability distribution of the undersampling pattern and its corresponding sampling rate. Further testing was performed using 3D T1-weighted MR images of the brain from the MICCAI 2013 Grand Challenge on Multi-Atlas Labeling dataset and locally acquired brain 3D T1-weighted MR images of healthy volunteers and contrast-enhanced 3D T1-weighted MR images of high-grade glioma patients. The results showed that the recovered MR images using our 2D probabilistic undersampling pattern (with or without the reconstruction network) significantly outperformed those using the existing start-of-the-art undersampling strategies for both qualitative and quantitative comparison, suggesting the advantages and some extent of the generalization of our proposed method.

Incidence of Developmental Venous Anomalies in Patients With Multiple Sclerosis: A 3 Tesla MRI Study.

ObjectivesThere is evidence of involvement of the venous system in multiple sclerosis (MS). If this bears also an association with the frequency and extent of developmental venous anomalies (DVA) still has to be determined. We therefore investigated this in patients with different phenotypes of MS and in comparison, to a control population.MethodsWe analyzed the contrast-enhanced T1-weighted MR scans of 431 patients (clinically isolated syndrome—CIS, n = 108; MS, n = 323) and of 162 control individuals for the presence of a DVA. We also measured the size of the DVA and draining vein and compared the DVA frequency between MS phenotypes.ResultsA DVA was found in 38 (8.8 %) of patients with CIS or MS and in 11 (6.8%) controls (p = 0.4). DVA frequency was highest in CIS (14.8%) and lowest in progressive MS (4.0%). The mean cranio-caudal and axial extension of the DVA was significantly lower in MS patients than controls (p < 0.05).ConclusionsThe frequency of DVA in MS patients is comparable to that in controls. Whether DVA size and appearance may change over time will have to be investigated in a longitudinal manner and with larger sample size.

SWI as an Alternative to Contrast-Enhanced Imaging to Detect Acute MS Lesions.

Acute inflammatory activity of MS lesions is traditionally assessed through contrast-enhanced T1-weighted MR images. The aim of our study was to determine whether a qualitative evaluation of non-contrast-enhanced SWI of new T2-hyperintense lesions might help distinguish acute and chronic lesions and whether it could be considered a possible alternative to gadolinium-based contrast agents for this purpose. Serial MR imaging studies from 55 patients with MS were reviewed to identify 169 new T2-hyperintense lesions. Two blinded neuroradiologists determined their signal pattern on SWI, considering 5 categories (hypointense rings, marked hypointensity, mild hypointensity, iso-/hyperintensity, indeterminate). Two different blinded neuroradiologists evaluated the presence or absence of enhancement in postcontrast T1-weighted images of the lesions. The Fisher exact test was used to determine whether each category of signal intensity on SWI was associated with gadolinium enhancement. The presence of hypointense rings or marked hypointensity showed a strong association with the absence of gadolinium enhancement (P < .001), with a sensitivity of 93.0% and a specificity of 82.9%. The presence of mild hypointensity or isohyperintensity showed a strong association with the presence of gadolinium enhancement (P < .001), with a sensitivity of 68.3% and a specificity of 99.2%. A qualitative analysis of the signal pattern on SWI of new T2-hyperintense MS lesions allows determining the likelihood that the lesions will enhance after administration of a gadolinium contrast agent, with high specificity albeit with a moderate sensitivity. While it cannot substitute for the use of contrast agent, it can be useful in some clinical settings in which the contrast agent cannot be administered.

Automated Koos Classification of Vestibular Schwannoma.

The Koos grading scale is a frequently used classification system for vestibular schwannoma (VS) that accounts for extrameatal tumor dimension and compression of the brain stem. We propose an artificial intelligence (AI) pipeline to fully automate the segmentation and Koos classification of VS from MRI to improve clinical workflow and facilitate patient management. We propose a method for Koos classification that does not only rely on available images but also on automatically generated segmentations. Artificial neural networks were trained and tested based on manual tumor segmentations and ground truth Koos grades of contrast-enhanced T1-weighted (ceT1) and high-resolution T2-weighted (hrT2) MR images from subjects with a single sporadic VS, acquired on a single scanner and with a standardized protocol. The first stage of the pipeline comprises a convolutional neural network (CNN) which can segment the VS and 7 adjacent structures. For the second stage, we propose two complementary approaches that are combined in an ensemble. The first approach applies a second CNN to the segmentation output to predict the Koos grade, the other approach extracts handcrafted features which are passed to a Random Forest classifier. The pipeline results were compared to those achieved by two neurosurgeons. Eligible patients (n = 308) were pseudo-randomly split into 5 groups to evaluate the model performance with 5-fold cross-validation. The weighted macro-averaged mean absolute error (MA-MAE), weighted macro-averaged F1 score (F1), and accuracy score of the ensemble model were assessed on the testing sets as follows: MA-MAE = 0.11 ± 0.05, F1 = 89.3 ± 3.0%, accuracy = 89.3 ± 2.9%, which was comparable to the average performance of two neurosurgeons: MA-MAE = 0.11 ± 0.08, F1 = 89.1 ± 5.2, accuracy = 88.6 ± 5.8%. Inter-rater reliability was assessed by calculating Fleiss' generalized kappa (k = 0.68) based on all 308 cases, and intra-rater reliabilities of annotator 1 (k = 0.95) and annotator 2 (k = 0.82) were calculated according to the weighted kappa metric with quadratic (Fleiss-Cohen) weights based on 15 randomly selected cases. We developed the first AI framework to automatically classify VS according to the Koos scale. The excellent results show that the accuracy of the framework is comparable to that of neurosurgeons and may therefore facilitate management of patients with VS. The models, code, and ground truth Koos grades for a subset of publicly available images (n = 188) will be released upon publication.

Automatic segmentation of head and neck primary tumors on MRI using a multi-view CNN

Background Accurate segmentation of head and neck squamous cell cancer (HNSCC) is important for radiotherapy treatment planning. Manual segmentation of these tumors is time-consuming and vulnerable to inconsistencies between experts, especially in the complex head and neck region. The aim of this study is to introduce and evaluate an automatic segmentation pipeline for HNSCC using a multi-view CNN (MV-CNN).MethodsThe dataset included 220 patients with primary HNSCC and availability of T1-weighted, STIR and optionally contrast-enhanced T1-weighted MR images together with a manual reference segmentation of the primary tumor by an expert. A T1-weighted standard space of the head and neck region was created to register all MRI sequences to. An MV-CNN was trained with these three MRI sequences and evaluated in terms of volumetric and spatial performance in a cross-validation by measuring intra-class correlation (ICC) and dice similarity score (DSC), respectively.ResultsThe average manual segmented primary tumor volume was 11.8±6.70 cm3 with a median [IQR] of 13.9 [3.22-15.9] cm3. The tumor volume measured by MV-CNN was 22.8±21.1 cm3 with a median [IQR] of 16.0 [8.24-31.1] cm3. Compared to the manual segmentations, the MV-CNN scored an average ICC of 0.64±0.06 and a DSC of 0.49±0.19. Improved segmentation performance was observed with increasing primary tumor volume: the smallest tumor volume group (<3 cm3) scored a DSC of 0.26±0.16 and the largest group (>15 cm3) a DSC of 0.63±0.11 (p<0.001). The automated segmentation tended to overestimate compared to the manual reference, both around the actual primary tumor and in false positively classified healthy structures and pathologically enlarged lymph nodes.ConclusionAn automatic segmentation pipeline was evaluated for primary HNSCC on MRI. The MV-CNN produced reasonable segmentation results, especially on large tumors, but overestimation decreased overall performance. In further research, the focus should be on decreasing false positives and make it valuable in treatment planning.

Sonobiopsy for minimally invasive, spatiotemporally-controlled, and sensitive detection of glioblastoma-derived circulating tumor DNA.

Though surgical biopsies provide direct access to tissue for genomic characterization of brain cancer, they are invasive and pose significant clinical risks. Brain cancer management via blood-based liquid biopsies is a minimally invasive alternative; however, the blood-brain barrier (BBB) restricts the release of brain tumor-derived molecular biomarkers necessary for sensitive diagnosis.Methods: A mouse glioblastoma multiforme (GBM) model was used to demonstrate the capability of focused ultrasound (FUS)-enabled liquid biopsy (sonobiopsy) to improve the diagnostic sensitivity of brain tumor-specific genetic mutations compared with conventional blood-based liquid biopsy. Furthermore, a pig GBM model was developed to characterize the translational implications of sonobiopsy in humans. Magnetic resonance imaging (MRI)-guided FUS sonication was performed in mice and pigs to locally enhance the BBB permeability of the GBM tumor. Contrast-enhanced T1-weighted MR images were acquired to evaluate the BBB permeability change. Blood was collected immediately after FUS sonication. Droplet digital PCR was used to quantify the levels of brain tumor-specific genetic mutations in the circulating tumor DNA (ctDNA). Histological staining was performed to evaluate the potential for off-target tissue damage by sonobiopsy.Results: Sonobiopsy improved the detection sensitivity of EGFRvIII from 7.14% to 64.71% and TERT C228T from 14.29% to 45.83% in the mouse GBM model. It also improved the diagnostic sensitivity of EGFRvIII from 28.57% to 100% and TERT C228T from 42.86% to 71.43% in the porcine GBM model.Conclusion: Sonobiopsy disrupts the BBB at the spatially-targeted brain location, releases tumor-derived DNA into the blood circulation, and enables timely collection of ctDNA. Converging evidence from both mouse and pig GBM models strongly supports the clinical translation of sonobiopsy for the minimally invasive, spatiotemporally-controlled, and sensitive molecular characterization of brain cancer.

Bone Metastases: State of the Art in Minimally Invasive Interventional Oncology.

Bone is the third most common site involved by cancer metastases, and skeleton-related events such as intractable pain due to direct osseous tumor involvement, pathologic fracture, and neurologic deficits as a consequence of nerve or spinal cord compression often affect patients' functional independence and quality of life unfavorably. The annual medical-economic burden related to bone metastases is a substantial component of the total direct medical cost estimated by the National Institutes of Health. There have been substantial recent advances in percutaneous image-guided minimally invasive musculoskeletal oncologic interventions for the management of patients with osseous metastatic disease. These advances include thermal ablation, cementation with or without osseous reinforcement with implants, osteosynthesis, thermal and chemical neurolyses, and palliative injections, which are progressively incorporated into the management paradigm for such patients. These interventions are performed in conjunction with or are supplemented by adjuvant radiation therapy, systemic therapy, surgery, or analgesic agents to achieve durable pain palliation, local tumor control, or cure, and they provide a robust armamentarium for interventional radiologists to achieve safe and effective treatment in a multidisciplinary setting. In addition, these procedures are shifting the patient management paradigm in modern-era practice. The authors detail the state of the art in minimally invasive percutaneous image-guided musculoskeletal oncologic interventions and the role of radiologists in managing patients with skeletal metastases. ©RSNA, 2021.

Ivy Sign in Moyamoya Disease: A Comparative Study of the FLAIR Vascular Hyperintensity Sign Against Contrast-Enhanced MRI.

The ability of the ivy sign on contrast-enhanced T1-weighted MR imaging (CEMR) to reflect cerebral perfusion and postoperative revascularization in Moyamoya disease remains largely unknown. We aimed to compare the capabilities of CEMR and FLAIR. CEMR, FLAIR, arterial spin-labeling, and DSA were performed in 44 patients with Moyamoya disease. The ivy sign was scored separately on CEMR and FLAIR using the Alberta Stroke Program Early CT Score. The status of leptomeningeal collaterals was scored on DSA. The postoperative Matsushima grade was evaluated at least 3 months after surgical revascularization. Scoring of the ivy sign on CEMR showed excellent interrater reliability, and FLAIR vascular hyperintensity showed moderate interrater reliability. Correlation analyses revealed that DSA scores were more consistent with the CEMR-based ivy sign score (r = 0.25, P = .03) than with FLAIR vascular hyperintensity (r = 0.05, P = .65). The CEMR-based ivy sign score was significantly correlated with CBF in late-Suzuki stage Moyamoya disease (t = -2.64, P = .02). The CEMR-based ivy sign score at baseline was significantly correlated with the postoperative Matsushima grade (r = 0.48, P = .03). In this study, CEMR outperformed FLAIR in capturing the ivy sign in Moyamoya disease. In addition, the CEMR-based ivy sign score provided adequate information on hemodynamic status and postoperative neovascularization. The current study suggested that CEMR could be considered as an alternative to FLAIR in future studies investigating leptomeningeal collaterals in Moyamoya disease.

Endoscope-assisted fluorescence-guided resection allowing supratotal removal in glioblastoma surgery.

Several studies have proven the benefits of a wide extent of resection (EOR) of contrast-enhancing tumor in terms of progression-free survival (PFS) and overall survival (OS) in patients with glioblastoma (GBM). Thus, gross-total resection (GTR) is the main surgical goal in noneloquently located GBMs. Complete tumor removal can be almost doubled by microscopic fluorescence guidance. Recently, a study has shown that an endoscope with a light source capable of inducing fluorescence allows visualization of remnant fluorescent tumor tissue even after complete microscopic fluorescence-guided (FG) resection, thereby increasing the rate of GTR. Since tumor infiltration spreads beyond the borders of contrast enhancement on MRI, the aim of this study was to determine via volumetric analyses of the EOR whether endoscope-assisted FG resection enables supratotal resection beyond the borders of contrast enhancement. The authors conducted a retrospective single-center analysis of a consecutive series of patients with primary GBM presumed to be noneloquently located and routinely operated on at their institution between January 2015 and February 2018 using a combined microscopic and endoscopic FG resection. A 20-mg/kg dose of 5-aminolevulinic acid (5-ALA) was administered 4 hours before surgery. After complete microscopic FG resection, the resection cavity was scanned using the endoscope. Detected residual fluorescent tissue was resected and embedded separately for histopathological examination. Nonenhanced and contrast-enhanced 3D T1-weighted MR images acquired before and within 48 hours after tumor resection were analyzed using 3D Slicer. Bias field-corrected data were used to segment brain parenchyma, contrast-enhancing tumor, and the resection cavity for volume definition. The difference between the pre- and postoperative brain parenchyma volume was considered to be equivalent to the resected nonenhancing but fluorescent tumor tissue. The volume of resected tumor tissue was calculated from the sum of resected contrast-enhancing tumor tissue and resected nonenhancing tumor tissue. Twelve patients with GBM were operated on using endoscopic after complete microscopic FG resection. In all cases, residual fluorescent tissue not visualized with the microscope was detected. Histopathological examination confirmed residual tumor tissue in all specimens. The mean preoperative volume of brain parenchyma without contrast-enhancing tumor was 1213.2 cm3. The mean postoperative volume of brain parenchyma without the resection cavity was 1151.2 cm3, accounting for a mean volume of nonenhancing but fluorescent tumor tissue of 62.0 cm3. The mean relative rate of the overall resected volume compared to the contrast-enhancing tumor volume was 244.7% (p < 0.001). Combined microscopic and endoscopic FG resection of GBM significantly increases the EOR and allows the surgeon to achieve a supratotal resection beyond the borders of contrast enhancement in noneloquently located GBM.