Search

Background Radiogenomics of pediatric medulloblastoma (MB) offers an opportunity for MB risk stratification, which may aid therapeutic decision making, family counseling, and selection of patient groups suitable for targeted genetic analysis. Purpose To develop machine learning strategies that identify the four clinically significant MB molecular subgroups. Materials and Methods In this retrospective study, consecutive pediatric patients with newly diagnosed MB at MRI at 12 international pediatric sites between July 1997 and May 2020 were identified. There were 1800 features extracted from T2- and contrast-enhanced T1-weighted preoperative MRI scans. A two-stage sequential classifier was designed-one that first identifies non-wingless (

A new task force dedicated to artificial intelligence (AI) with respect to paediatric radiology was created in 2021 at the International Paediatric Radiology (IPR) meeting in Rome, Italy (a joint society meeting by the European Society of Pediatric Radiology [ESPR] and the Society for Pediatric Radiology [SPR]). The concept of a separate task force dedicated to AI was borne from an ESPR-led international survey of health care professionals’ opinions, expectations and concerns regarding AI integration within children’s imaging departments. In this survey, the majority (> 80%) of ESPR respondents supported the creation of a task force and helped define our key objectives. These include providing educational content about AI relevant for paediatric radiologists, brainstorming ideas for future projects and collaborating on AI-related studies with respect to collating data sets, de-identifying images and engaging in multi-case, multi-reader studies. This manuscript outlines the starting point of the ESPR AI task force and where we wish to go.

Background The risk profile for posterior fossa ependymoma (EP) depends on surgical and molecular status [Group A (PFA) versus Group B (PFB)]. While subtotal tumor resection is known to confer worse prognosis, MRI-based EP risk-profiling is unexplored. We aimed to apply machine learning strategies to link MRI-based biomarkers of high-risk EP and also to distinguish PFA from PFB. Methods We extracted 1800 quantitative features from presurgical T2-weighted (T2-MRI) and gadolinium-enhanced T1-weighted (T1-MRI) imaging of 157 EP patients. We implemented nested cross-validation to identify features for risk score calculations and apply a Cox model for survival analysis. We conducted additional feature selection for PFA versus PFB and examined performance across three candidate classifiers. Results For all EP patients with GTR, we identified four T2-MRI-based features and stratified patients into high- and low-risk groups, with 5-year overall survival rates of 62% and 100%, respectively (P < .0001). Among presumed PFA patients with GTR, four T1-MRI and five T2-MRI features predicted divergence of high- and low-risk groups, with 5-year overall survival rates of 62.7% and 96.7%, respectively (P = .002). T1-MRI-based features showed the best performance distinguishing PFA from PFB with an AUC of 0.86. Conclusions We present machine learning strategies to identify MRI phenotypes that distinguish PFA from PFB, as well as high- and low-risk PFA. We also describe quantitative image predictors of aggressive EP tumors that might assist risk-profiling after surgery. Future studies could examine translating radiomics as an adjunct to EP risk assessment when considering therapy strategies or trial candidacy.

Background and purpose Recurrent seizures have been reported to induce neuronal loss in the hippocampus. It is unclear whether seizure control influences hippocampal volume. The aims of this study were to determine if there was a change in total or subfield hippocampal volume over time in children with focal drug-resistant epilepsy, and whether seizure control influenced total or subfield hippocampal volumes. Methods Using FreeSurfer’s automated segmentation of brain magnetic resonance imaging scans, we calculated the total and subfield (including CA1, CA3, CA4, subiculum, presubiculum, parasubiculum, molecular layer and dentate gyrus) hippocampal volumes of children with non-lesional focal epilepsy. Seizure frequency and hippocampal volumes were assessed at baseline and follow-up. Patients were classified into those who were seizure free or have improvement in seizures (group 1) and those with no improvement in seizures (group 2) at follow-up. Results Thirty-seven patients were included, with mean age 10.31 ± 3.68 years at baseline. The interval between the two magnetic resonance imaging scans was 2.59 ± 1.25 years. There was no significant difference in the total and subfield hippocampal volumes for the whole cohort at follow-up compared to baseline (all P > 0.002). Seizure control of the two groups did not predict total or subfield hippocampal volume, after controlling for baseline volume, age, severity of seizure frequency at baseline and time interval between the magnetic resonance imaging scans (all P > 0.002). Conclusion We have found that total and subfield hippocampal volumes did not change, and seizure control did not predict hippocampal volumes at follow-up in children with drug-resistant epilepsy.

Purpose Artificial intelligence (AI) is playing an ever-increasing role in Neuroradiology. Methods When designing AI-based research in neuroradiology and appreciating the literature, it is important to understand the fundamental principles of AI. Training, validation, and test datasets must be defined and set apart as priorities. External validation and testing datasets are preferable, when feasible. The specific type of learning process (supervised vs. unsupervised) and the machine learning model also require definition. Deep learning (DL) is an AI-based approach that is modelled on the structure of neurons of the brain; convolutional neural networks (CNN) are a commonly used example in neuroradiology. Results Radiomics is a frequently used approach in which a multitude of imaging features are extracted from a region of interest and subsequently reduced and selected to convey diagnostic or prognostic information. Deep radiomics uses CNNs to directly extract features and obviate the need for predefined features. Conclusion Common limitations and pitfalls in AI-based research in neuroradiology are limited sample sizes (“small-n-large-p problem”), selection bias, as well as overfitting and underfitting.

Background Clinicians and machine classifiers reliably diagnose pilocytic astrocytoma (PA) on magnetic resonance imaging (MRI) but less accurately distinguish medulloblastoma (MB) from ependymoma (EP). One strategy is to first rule out the most identifiable diagnosis. Objective To hypothesize a sequential machine-learning classifier could improve diagnostic performance by mimicking a clinician's strategy of excluding PA before distinguishing MB from EP. Methods We extracted 1800 total Image Biomarker Standardization Initiative (IBSI)-based features from T2- and gadolinium-enhanced T1-weighted images in a multinational cohort of 274 MB, 156 PA, and 97 EP. We designed a 2-step sequential classifier - first ruling out PA, and next distinguishing MB from EP. For each step, we selected the best performing model from 6-candidate classifier using a reduced feature set, and measured performance on a holdout test set with the microaveraged F1 score. Results Optimal diagnostic performance was achieved using 2 decision steps, each with its own distinct imaging features and classifier method. A 3-way logistic regression classifier first distinguished PA from non-PA, with T2 uniformity and T1 contrast as the most relevant IBSI features (F1 score 0.8809). A 2-way neural net classifier next distinguished MB from EP, with T2 sphericity and T1 flatness as most relevant (F1 score 0.9189). The combined, sequential classifier was with F1 score 0.9179. Conclusion An MRI-based sequential machine-learning classifiers offer high-performance prediction of pediatric posterior fossa tumors across a large, multinational cohort. Optimization of this model with demographic, clinical, imaging, and molecular predictors could provide significant advantages for family counseling and surgical planning.

In the past decade, a new approach for quantitative analysis of medical images and prognostic modelling has emerged. Defined as the extraction and analysis of a large number of quantitative parameters from medical images, radiomics is an evolving field in precision medicine with the ultimate goal of the discovery of new imaging biomarkers for disease. Radiomics has already shown promising results in extracting diagnostic, prognostic, and molecular information latent in medical images. After acquisition of the medical images as part of the standard of care, a region of interest is defined often via a manual or semi-automatic approach. An algorithm then extracts and computes quantitative radiomics parameters from the region of interest. Whereas radiomics captures quantitative values of shape and texture based on predefined mathematical terms, neural networks have recently been used to directly learn and identify predictive features from medical images. Thereby, neural networks largely forego the need for so called "hand-engineered" features, which appears to result in significantly improved performance and reliability. Opportunities for radiomics and neural networks in pediatric nuclear medicine/radiology/molecular imaging are broad and can be thought of in three categories: automating well-defined administrative or clinical tasks, augmenting broader administrative or clinical tasks, and unlocking new methods of generating value. Specific applications include intelligent order sets, automated protocoling, improved image acquisition, computer aided triage and detection of abnormalities, next generation voice dictation systems, biomarker development, and therapy planning.

Purpose: To assess the accuracy of answers provided by ChatGPT-3 when prompted with questions from the daily routine of radiologists and to evaluate the text response when ChatGPT-3 was prompted to provide references for a given answer. Methods: ChatGPT-3 (San Francisco, OpenAI) is an artificial intelligence chatbot based on a large language model (LLM) that has been designed to generate human-like text. A total of 88 questions were submitted to ChatGPT-3 using textual prompt. These 88 questions were equally dispersed across 8 subspecialty areas of radiology. The responses provided by ChatGPT-3 were assessed for correctness by cross-checking them with peer-reviewed, PubMed-listed references. In addition, the references provided by ChatGPT-3 were evaluated for authenticity. Results: A total of 59 of 88 responses (67%) to radiological questions were correct, while 29 responses (33%) had errors. Out of 343 references provided, only 124 references (36.2%) were available through internet search, while 219 references (63.8%) appeared to be generated by ChatGPT-3. When examining the 124 identified references, only 47 references (37.9%) were considered to provide enough background to correctly answer 24 questions (37.5%). Conclusion: In this pilot study, ChatGPT-3 provided correct responses to questions from the daily clinical routine of radiologists in only about two thirds, while the remainder of responses contained errors. The majority of provided references were not found and only a minority of the provided references contained the correct information to answer the question. Caution is advised when using ChatGPT-3 to retrieve radiological information.

Objectives Histological studies have shown alterations of thalamic nuclei in patients with Down syndrome (DS). The correlation of these changes on MRI (magnetic resonance imaging) is unclear. Therefore, this study investigates volumetric differences of thalamic nuclei in children with DS compared to controls. Methods Patients were retrospectively identified between 01/2000 and 10/2021. Patient inclusion criteria were: (1) 0–18 years of age, (2) diagnosis of DS, and (3) availability of a brain MRI without parenchymal injury and a non-motion-degraded volumetric T1-weighted sequence. Whole thalamus and thalamic nuclei ( n = 25) volumes were analyzed bilaterally relative to the total brain volume (TBV). Two-sided t-tests were used to evaluate differences between groups. Differences were considered significant if the adjusted p-value was Results 21 children with DS (11 females, 52.4%, mean age: 8.6 ± 4.3 years) and 63 age- and sex-matched controls (32 females, 50.8%, 8.6 ± 4.3 years) were studied using automated volumetric segmentation. Significantly smaller ratios were found for nine thalamic nuclei and the whole thalamus on the right and five thalamic nuclei on the left. TBV was significantly smaller in patients with DS ( p < 0.001). No significant differences were found between the groups for age and sex. Conclusions In this exploratory volumetric analysis of the thalamus and thalamic nuclei, we observed statistically significant volumetric changes in children with DS. Our findings confirm prior neuroimaging and histological studies and extend the range of involved thalamic nuclei in pediatric DS.

OBJECTIVEIn comparison with the general population, children with syndromic craniosynostosis (sCS) have abnormal cerebral venous anatomy and are more likely to develop intracranial hypertension. To date, little is known about the postnatal development change in cerebral blood flow (CBF) in sCS. The aim of this study was to determine CBF in patients with sCS, and compare findings with control subjects.METHODSA prospective cohort study of patients with sCS using MRI and arterial spin labeling (ASL) determined regional CBF patterns in comparison with a convenience sample of control subjects with identical MRI/ASL assessments in whom the imaging showed no cerebral/neurological pathology. Patients with SCS and control subjects were stratified into four age categories and compared using CBF measurements from four brain lobes, the cerebellum, supratentorial cortex, and white matter. In a subgroup of patients with sCS the authors also compared longitudinal pre- to postoperative CBF changes.RESULTSSeventy-six patients with sCS (35 female [46.1%] and 41 male [53.9%]), with a mean age of 4.5 years (range 0.2–19.2 years), were compared with 86 control subjects (38 female [44.2%] and 48 male [55.8%]), with a mean age of 6.4 years (range 0.1–17.8 years). Untreated sCS patients < 1 year old had lower CBF than control subjects. In older age categories, CBF normalized to values observed in controls. Graphical analyses of CBF by age showed that the normally expected peak in CBF during childhood, noted at 4 years of age in control subjects, occurred at 5–6 years of age in patients with sCS. Patients with longitudinal pre- to postoperative CBF measurements showed significant increases in CBF after surgery.CONCLUSIONSUntreated patients with sCS < 1 year old have lower CBF than control subjects. Following vault expansion, and with age, CBF in these patients normalizes to that of control subjects, but the usual physiological peak in CBF in childhood occurs later than expected.

The integration of human and machine intelligence promises to profoundly change the practice of medicine. The rapidly increasing adoption of artificial intelligence (AI) solutions highlights its potential to streamline physician work and optimize clinical decision-making, also in the field of pediatric radiology. Large imaging databases are necessary for training, validating and testing these algorithms. To better promote data accessibility in multi-institutional AI-enabled radiologic research, these databases centralize the large volumes of data required to effect accurate models and outcome predictions. However, such undertakings must consider the sensitivity of patient information and therefore utilize requisite data governance measures to safeguard data privacy and security, to recognize and mitigate the effects of bias and to promote ethical use. In this article we define data stewardship and data governance, review their key considerations and applicability to radiologic research in the pediatric context, and consider the associated best practices along with the ramifications of poorly executed data governance. We summarize several adaptable data governance frameworks and describe strategies for their implementation in the form of distributed and centralized approaches to data management.

Deep learning techniques using convolutional neural networks (CNNs) have been successfully developed for various medical image analysis tasks. However, the skills to understand and develop deep learning models are not usually taught during radiology training, which constitutes a barrier for radiologists looking to integrate machine learning (ML) into their research or clinical practice. In this work, we developed and evaluated an educational graphical user interface (GUI) to construct CNNs for teaching deep learning concepts to radiology trainees. The GUI was developed in Python using the PyQt and PyTorch frameworks. The functionality of the GUI was demonstrated through a binary classification task on a dataset of MR images of the brain. The usability of the GUI was assessed through 45-min user testing sessions with 5 neuroradiologists and neuroradiology fellows, assessing mean task completion times, the System Usability Scale (SUS), and a qualitative questionnaire as metrics. Task completion times were compared against a ML expert who performed the same tasks. After a 20-min introduction to CNNs and a walkthrough of the GUI, users were able to perform all assigned tasks successfully. There was no significant difference in task completion time compared to a ML expert. The educational GUI achieved a score of 82.5 on the SUS, suggesting that the system is highly usable. Users indicated that the GUI seems useful as an educational tool to teach ML topics to radiology trainees. An educational GUI allows interactive teaching in ML that can be incorporated into radiology training.

Machine learning (ML) approaches can predict BRAF status of pediatric low-grade gliomas (pLGG) on pre-therapeutic brain MRI. The impact of training data sample size and type of ML model is not established. In this bi-institutional retrospective study, 251 pLGG FLAIR MRI datasets from 2 children’s hospitals were included. Radiomics features were extracted from tumor segmentations and five models (Random Forest, XGBoost, Neural Network (NN) 1 (100:20:2), NN2 (50:10:2), NN3 (50:20:10:2)) were tested to classify them. Classifiers were cross-validated on data from institution 1 and validated on data from institution 2. Starting with 10% of the training data, models were cross-validated using a 4-fold approach at every step with an additional 2.25% increase in sample size. Two-hundred-twenty patients (mean age 8.53 ± 4.94 years, 114 males, 67% BRAF fusion) were included in the training dataset, and 31 patients (mean age 7.97±6.20 years, 18 males, 77% BRAF fusion) in the independent test dataset. NN1 (100:20:2) yielded the highest area under the receiver operating characteristic curve (AUC). It predicted BRAF status with a mean AUC of 0.85, 95% CI [0.83, 0.87] using 60% of the training data and with mean AUC of 0.83, 95% CI [0.82, 0.84] on the independent validation data set.

Segmentation of the fetus from 2-dimensional (2D) magnetic resonance imaging (MRI) can aid radiologists with clinical decision making for disease diagnosis. Machine learning can facilitate this process of automatic segmentation, making diagnosis more accurate and user independent. We propose a deep learning (DL) framework for 2D fetal MRI segmentation using a Cross Attention Squeeze Excitation Network (CASE-Net) for research and clinical applications. CASE-Net is an end-to-end segmentation architecture with relevant modules that are evidence based. The goal of CASE-Net is to emphasize localization of contextual information that is relevant in biomedical segmentation, by combining attention mechanisms with squeeze-and-excitation (SE) blocks. This is a retrospective study with 34 patients. Our experiments have shown that our proposed CASE-Net achieved the highest segmentation Dice score of 87.36%, outperforming other competitive segmentation architectures.

Radiation-associated aneurysms are rare, difficult to treat, and associated with high morbidity and mortality when ruptured, compared with aneurysms unrelated to radiation treatment. We present a 16-year-old patient with a radiation-induced intracranial aneurysm arising from the left posterior inferior cerebellar artery (PICA), 10 years following radiotherapy for medulloblastoma. The patient successfully underwent endovascular coil embolization of the parent artery across the neck of the aneurysm. CT angiography and MRI in the days following the procedure demonstrated maintained flow in the anterior and lateral medullary PICA segments with no brainstem infarct.

OBJECTIVE While intraventricular hemorrhage (IVH) is associated with posthemorrhagic ventricular dilation (PHVD), not all infants affected by high-grade IVH develop PHVD. The authors aimed to determine clot-associated predictors of PHVD in a porcine model by varying the amount and rate of direct intraventricular injection of whole autologous blood. METHODS Seven 1-week-old piglets underwent craniectomy and injection of autologous blood into the right lateral ventricle. They survived for a maximum of 28 days. MRI was performed prior to injection, immediately postoperatively, and every 7 days thereafter. T1-weighted, T2-weighted, and susceptibility-weighted imaging (SWI) sequences were used to segment ventricular and clot volumes. Spearman correlations were used to determine the relationship between blood and clot volumes and ventricular volumes over time. RESULTS The maximum ventricular volume was up to 12 times that of baseline. One animal developed acute hydrocephalus on day 4. All other animals survived until planned endpoints. The interaction between volume of blood injected and duration of injection was significantly associated with clot volume on the postoperative scan (p = 0.003) but not the amount of blood injected alone (p = 0.38). Initial postoperative and day 7 clot volumes, but not volume of blood injected, were correlated with maximum (p = 0.007 and 0.014) and terminal (p = 0.014 and 0.036) ventricular volumes. Initial postoperative ventricular volume was correlated with maximum and terminal ventricular volume (p = 0.007 and p = 0.014). CONCLUSIONS Initial postoperative, maximum, and terminal ventricular dilations were associated with the amount of clot formed, rather than the amount of blood injected. This supports the hypothesis that PHVD is determined by clot burden rather than the presence of blood products and allows further testing of early clot lysis to minimize PHVD risk.

Low-dose CT is the modality of choice in the evaluation of craniosynostosis, and use of an essential imaging protocol and knowledge of syndromic causes are helpful in preoperative assessment.

Graphical abstract, Highlights • MRI is the modality of choice to image and quantify cerebral perfusion. • Imaging of neonatal brain perfusion is possible using MRI and ultrasound. • Novel ultrafast ultrasound imaging allows for excellent spatiotemporal resolution. • Understanding cerebral hemodynamic changes of neonatal adaptation is key., Abnormal variations of the neonatal brain perfusion can result in long-term neurodevelopmental consequences and cerebral perfusion imaging can play an important role in diagnostic and therapeutic decision-making. To identify at-risk situations, perfusion imaging of the neonatal brain must accurately evaluate both regional and global perfusion. To date, neonatal cerebral perfusion assessment remains challenging. The available modalities such as magnetic resonance imaging (MRI), ultrasound imaging, computed tomography (CT), near-infrared spectroscopy or nuclear imaging have multiple compromises and limitations. Several promising methods are being developed to achieve better diagnostic accuracy and higher robustness, in particular using advanced MRI and ultrasound techniques. The objective of this state-of-the-art review is to analyze the methodology and challenges of neonatal brain perfusion imaging, to describe the currently available modalities, and to outline future perspectives.

Radiation-associated aneurysms are rare, difficult to treat, and associated with high morbidity and mortality when ruptured, compared with aneurysms unrelated to radiation treatment. We present a 16-year-old patient with a radiation-induced intracranial aneurysm arising from the left posterior inferior cerebellar artery (PICA), 10 years following radiotherapy for medulloblastoma. The patient successfully underwent endovascular coil embolization of the parent artery across the neck of the aneurysm. CT angiography and MRI in the days following the procedure demonstrated maintained flow in the anterior and lateral medullary PICA segments with no brainstem infarct.

Genetic skeletal disorders (GSDs) are a heterogeneous group characterized by an intrinsic abnormality in growth and (re-)modeling of cartilage and bone. A large subgroup of GSDs has additional involvement of other structures/organs beside the skeleton, such as the central nervous system (CNS). CNS abnormalities have an important role in long-term prognosis of children with GSDs and should consequently not be missed. Sensitive and specific identification of CNS lesions while evaluating a child with a GSD requires a detailed knowledge of the possible associated CNS abnormalities. Here, we provide a pattern-recognition approach for neuroimaging findings in GSDs guided by the obvious skeletal manifestations of GSD. In particular, we summarize which CNS findings should be ruled out with each GSD. The diseases (n = 180) are classified based on the skeletal involvement (1. abnormal metaphysis or epiphysis, 2. abnormal size/number of bones, 3. abnormal shape of bones and joints, and 4. abnormal dynamic or structural changes). For each disease, skeletal involvement was defined in accordance with Online Mendelian Inheritance in Man. Morphological CNS involvement has been described based on extensive literature search. Selected examples will be shown based on prevalence of the diseases and significance of the CNS involvement. CNS involvement is common in GSDs. A wide spectrum of morphological abnormalities is associated with GSDs. Early diagnosis of CNS involvement is important in the management of children with GSDs. This pattern-recognition approach aims to assist and guide physicians in the diagnostic work-up of CNS involvement in children with GSDs and their management.

Background Our aims were to determine if features derived from texture analysis (TA) can distinguish normal, benign, and malignant tissue on automated breast ultrasound (ABUS); to evaluate whether machine learning (ML) applied to TA can categorise ABUS findings; and to compare ML to the analysis of single texture features for lesion classification. Methods This ethically approved retrospective pilot study included 54 women with benign (n = 38) and malignant (n = 32) solid breast lesions who underwent ABUS. After manual region of interest placement along the lesions’ margin as well as the surrounding fat and glandular breast tissue, 47 texture features (TFs) were calculated for each category. Statistical analysis (ANOVA) and a support vector machine (SVM) algorithm were applied to the texture feature to evaluate the accuracy in distinguishing (i) lesions versus normal tissue and (ii) benign versus malignant lesions. Results Skewness and kurtosis were the only TF significantly different among all the four categories (p < 0.000001). In subsets (i) and (ii), a maximum area under the curve of 0.86 (95% confidence interval [CI] 0.82–0.88) for energy and 0.86 (95% CI 0.82–0.89) for entropy were obtained. Using the SVM algorithm, a maximum area under the curve of 0.98 for both subsets was obtained with a maximum accuracy of 94.4% in subset (i) and 90.7% in subset (ii). Conclusions TA in combination with ML might represent a useful diagnostic tool in the evaluation of breast imaging findings in ABUS. Applying ML techniques to TFs might be superior compared to the analysis of single TF.

BACKGROUND AND PURPOSE To determine the feasibility of apparent diffusion coefficient (ADC)-based thermometry to assess intraventricular temperature in children. METHODS ADC maps were generated from diffusion tensor imaging data, which were acquired with diffusion gradients along 20 noncollinear directions using a b-value of 1000 s/mm2. The intraventricular temperature was calculated based on intraventricular ADC values and the mode method as previously reported. The calculated intraventricular temperature was validated with an estimated brain temperature based on temporal artery temperature measurements. We included 120 children in this study (49 females, 71 males, mean age 6.63 years), 15 consecutive children for each of the following age groups: 0-1, 1-2, 2-4, 4-6, 6-8, 8-10, 10-14, and 14-18 years. Forty-three children had a normal brain MRI and 77 children had an abnormal brain scan. Polynomial fitting to the temperature distribution and subsequent calculation of mode values was performed. A correlation coefficient and a coefficient of determination were calculated between ADC calculated temperatures and estimated brain temperatures. Linear regression analysis was performed to investigate the two temperature measures. RESULTS ADC-based intraventricular temperatures ranged between 31.5 and 39.6 °C, although estimated brain temperatures ranged between 36.3 and 38.1 °C. The difference between the temperatures is larger for children with more than 8,000 voxels within the lateral ventricles compared to children with less than 8,000 voxels. The correlation coefficient between ADC-based temperatures and the estimated brain temperatures is .1, the respective R2 is .01 indicating that 1% of the changes in estimated brain temperatures are attributable to corresponding changes in ADC-based temperature measurements (P = .275). CONCLUSIONS ADC-based thermometry has limited application in the pediatric population mainly due to a small ventricular size.

The aim of this study was to assess the interreader agreement and diagnostic accuracy of morphologic magnetic resonance imaging (MRI) analysis and quantitative MRI-based texture analysis (TA) for grading of cartilaginous bone tumors.This retrospective study was approved by our local ethics committee. Magnetic resonance imaging scans of 116 cartilaginous bone neoplasms were included (53 chondromas, 26 low-grade chondrosarcomas, 37 high-grade chondrosarcomas). Two musculoskeletal radiologists blinded to patient data separately analyzed 14 morphologic MRI features consisting of tumor and peritumoral characteristics. In addition, 2 different musculoskeletal radiologists separately performed TA including 19 quantitative TA parameters in a similar fashion. Interreader reliability, univariate, multivariate, and receiver operating characteristics analyses were performed for MRI and TA parameters separately and for combined models to determine independent predictors and diagnostic accuracy for grading of cartilaginous neoplasms. P values of 0.05 and less were considered statistically significant.Between both readers, MRI and TA features showed a mean kappa value of 0.49 (range, 0.08-0.82) and a mean intraclass correlation coefficient of 0.79 (range, 0.43-0.99), respectively. Independent morphological MRI predictors for grading of cartilaginous neoplasms were bone marrow edema, soft tissue mass, maximum tumor extent, and active periostitis, whereas TA predictors consisted of short-run high gray-level emphasis, skewness, and gray-level and run-length nonuniformity. Diagnostic accuracies for differentiation of benign from malignant as well as for benign from low-grade cartilaginous lesions were 87.0% and 77.4% using MRI predictors exclusively, 89.8% and 89.5% using TA predictors exclusively, and 92.9% and 91.2% using a combined model of MRI and TA predictors, respectively. For differentiation of low-grade from high-grade chondrosarcoma, no statistically significant independent TA predictors existed, whereas a model containing MRI predictors exclusively had a diagnostic accuracy of 84.8%.Texture analysis improves diagnostic accuracy for differentiation of benign and malignant as well as for benign and low-grade cartilaginous lesions when compared with morphologic MRI analysis.

Background The purpose of this study was to investigate whether any texture features show a correlation with intrahepatic tumor growth before the metastasis is visible to the human eye. Methods Eight male C57BL6 mice (age 8–10 weeks) were injected intraportally with syngeneic MC-38 colon cancer cells and two mice were injected with phosphate-buffered saline (sham controls). Small animal magnetic resonance imaging (MRI) at 4.7 T was performed at baseline and days 4, 8, 12, 16, and 20 after injection applying a T2-weighted spin-echo sequence. Texture analysis was performed on the images yielding 32 texture features derived from histogram, gray-level co-occurrence matrix, gray-level run-length matrix, and gray-level size-zone matrix. The features were examined with a linear regression model/Pearson correlation test and hierarchical cluster analysis. From each cluster, the feature with the lowest variance was selected. Results Tumors were visible on MRI after 20 days. Eighteen features from histogram and the gray-level-matrices exhibited statistically significant correlations before day 20 in the experiment group, but not in the control animals. Cluster analysis revealed three distinct clusters of independent features. The features with the lowest variance were Energy, Short Run Emphasis, and Gray Level Non-Uniformity. Conclusions Texture features may quantitatively detect liver metastases before they become visually detectable by the radiologist.

BACKGROUND AND PURPOSE To determine if axial T2-weighted imaging can serve as screening tool for pediatric brain imaging. METHODS We retrospectively evaluated consecutive brain magnetic resonance imaging (MRI) data of 161 children (74 girls) with a mean age of 7.44 ± 5.71 years. Standard of reference was the final report of neuroradiology attendings. Three readers with different levels of experience were blinded for clinical diagnoses and study indications. First, readers studied only the axial T2-weighted screening sequence. Second, they studied all available anatomical and functional MRI sequences as performed per standard protocol for each clinical indication. The readings were classified as normal or abnormal. Sensitivity and specificity were measured. RESULTS Axial T2 screening yielded a sensitivity of 77-88% and a specificity of 92%. The full studies/data sets had a sensitivity of 89-95% and a specificity of 86-93%. Nineteen of 167 studies were acquired for acute and 148 of 167 studies for nonacute clinical indication. Twenty-five false-negative diagnoses paneled in three groups were made by all readers together. Readers misread four of 19 studies with acute and 21 of 148 studies with nonacute clinical indication. Four of 21 misread studies with nonacute indications harbored unexpected findings needing management. CONCLUSIONS Axial T2 screening can detect pediatric brain abnormalities with high sensitivity and specificity and can possibly replace CT as screening tool if the reading physician is aware of possible limitations/pitfalls. The level of experience influences sensitivity and specificity. Adding diffusion-weighted imaging and susceptibility-weighted imaging to a 3-dimensional T2-weighted sequence would most likely further increase sensitivity and specificity.

BACKGROUND AND PURPOSE Apparent diffusion coefficient (ADC) values have been shown to assist in differentiating cerebellar pilocytic astrocytomas and medulloblastomas. Previous studies have applied only ADC measurements and calculated the mean/median values. Here we investigated the value of diffusion tensor imaging (DTI) histogram characteristics of the entire tumor for differentiation of cerebellar pilocytic astrocytomas and medulloblastomas. METHODS Presurgical DTI data were analyzed with a region of interest (ROI) approach to include the entire tumor. For each tumor, histogram-derived metrics including the 25th percentile, 75th percentile, and skewness were calculated for fractional anisotropy (FA) and mean (MD), axial (AD), and radial (RD) diffusivity. The histogram metrics were used as primary predictors of interest in a logistic regression model. Statistical significance levels were set at p < .01. RESULTS The study population included 17 children with pilocytic astrocytoma and 16 with medulloblastoma (mean age, 9.21 ± 5.18 years and 7.66 ± 4.97 years, respectively). Compared to children with medulloblastoma, children with pilocytic astrocytoma showed higher MD (P = .003 and P = .008), AD (P = .004 and P = .007), and RD (P = .003 and P = .009) values for the 25th and 75th percentile. In addition, histogram skewness showed statistically significant differences for MD between low- and high-grade tumors (P = .008). CONCLUSIONS The 25th percentile for MD yields the best results for the presurgical differentiation between pediatric cerebellar pilocytic astrocytomas and medulloblastomas. The analysis of other DTI metrics does not provide additional diagnostic value. Our study confirms the diagnostic value of the quantitative histogram analysis of DTI data in pediatric neuro-oncology.

Pediatric cerebral sinovenous thrombosis (CSVT) is a potentially life-threatening condition which is usually diagnosed by MRI. We analyzed the signal changes of the thrombus over time and the role of diffusion-weighted/tensor imaging (DWI/DTI) in the diagnosis of CSVT. Clinical histories were reviewed for risk factors for CSVT, neurologic manifestation, and interval from onset of symptoms related to CSVT to the neuroimaging diagnosis. MRI studies were retrospectively evaluated for the appearance of thrombi on T1- and T2-weighted, fluid-attenuated inversion recovery (FLAIR), DWI/DTI, susceptibility-weighted imaging (SWI), and magnetic resonance venography (MRV) images. Thirty-three children with CSVT were included in this study. Seventy-seven thrombi were found. Seventy-four thrombi could be identified on T1- or T2-weighted images (96 %), 72 thrombi were seen on DWI/DTI (94 %) and 68 on FLAIR (88 %). DWI showed restricted diffusion in 29 thrombi (40 %). Thrombi older than 1 day were more likely to have a T1-hyperintense signal (p = 0.002). No additional correlation between signal intensity and age of the thrombi was found. Intraparenchymal changes secondary to CSVT were seen in 11 children. MR sequences individually are not sensitive enough to provide the diagnosis. DWI/DTI does not provide complementary diagnostic value. Approximation of the age of the thrombus is difficult because of poor correlation between signal intensity and age of the thrombi.

Pediatric spinal trauma is unique. The developing pediatric spinal column and spinal cord deal with direct impact and indirect acceleration/deceleration or shear forces very different compared to adult patients. In addition children are exposed to different kind of traumas. Moreover, each age group has its unique patterns of injury. Familiarity with the normal developing spinal anatomy and kind of traumas is essential to correctly diagnose injury. Various imaging modalities can be used. Ultrasound is limited to the neonatal time period; plain radiography and computer tomography are typically used in the acute work-up and give highly detailed information about the osseous lesions. Magnetic resonance imaging is more sensitive for disco-ligamentous and spinal cord injuries. Depending on the clinical presentation and timing of trauma the various imaging modalities will be employed. In the current review article, a summary of the epidemiology and distribution of posttraumatic lesions is discussed in the context of the normal anatomical variations due to progressing development of the child.

Background Texture analysis in oncological magnetic resonance imaging (MRI) may yield surrogate markers for tumor differentiation and staging, both of which are important factors in the treatment planning for cervical cancer. Purpose To identify texture features which may predict tumor differentiation and nodal status in diffusion-weighted imaging (DWI) of cervical carcinoma Material and Methods Twenty-three patients were enrolled in this prospective, institutional review board (IRB)-approved study. Pelvic MRI was performed at 3-T including a DWI echo-planar sequence with b-values 40, 300, and 800 s/mm2. Apparent diffusion coefficient (ADC) maps were used for region of interest (ROI)-based texture analysis (32 texture features) of tumor, muscle, and fat based on histogram and gray-level matrices (GLM). All features confounded by the ROI size (linear model) were excluded. The remaining features were examined for correlations with histological differentiation (Spearman) and nodal status (Kruskal–Wallis). Hierarchical cluster analysis was used to identify correlations between features. A P value Results Mean age was 55 years (range = 37–78 years). Biopsy revealed two well-differentiated, eight moderately differentiated, two moderately to poorly differentiated tumors, and five poorly differentiated tumors. Six tumors could not be graded. Lymph nodes were involved in 11 patients. Three GLM features correlated with the differentiation: LRHGE (ϱ = 0.53, P = 0.03), ZP (ϱ = –0.49, P Conclusion We identified potentially predictive GLM features for histological tumor differentiation and histogram features for nodal cancer stage.

BACKGROUND AND PURPOSE: Head CT is the current neuroimaging tool of choice in acute evaluation of pediatric head trauma. The potential cancer risks of CT-related ionizing radiation should limit its use in children. We evaluated the role of MR imaging, including a “black bone” sequence, compared with CT in detecting skull fractures and intracranial hemorrhages in children with acute head trauma. MATERIALS AND METHODS: We performed a retrospective evaluation of 2D head CT and brain MR imaging studies including the black bone sequence of children with head trauma. Two experienced pediatric neuroradiologists in consensus created the standard of reference. Another pediatric neuroradiologist blinded to the diagnosis evaluated brain MR images and head CT images in 2 separate sessions. The presence of skull fractures and intracranial posttraumatic hemorrhages was evaluated. We calculated the sensitivity and specificity of CT and MR imaging with the black bone sequence in the diagnosis of skull fractures and intracranial hemorrhages. RESULTS: Twenty-eight children (24 boys; mean age, 4.89 years; range, 0–15.5 years) with head trauma were included. MR imaging with the black bone sequence revealed lower sensitivity (66.7% versus 100%) and specificity (87.5% versus 100%) in identifying skull fractures. Four of 6 incorrectly interpreted black bone MR imaging studies showed cranial sutures being misinterpreted as skull fractures and vice versa. CONCLUSIONS: Our preliminary results show that brain MR imaging complemented by a black bone sequence is a promising nonionizing alternative to head CT for the assessment of skull fractures in children. However, accuracy in the detection of linear fractures in young children and fractures of aerated bone remains limited.

Dermoid cysts are benign epithelial inclusions and cystic lesions that may occur in several locations including the oropharynx. We describe the case of a two-month-old baby girl who presented with progressive respiratory distress, hypoxemia, and feeding difficulties because of an oropharyngeal dermoid cyst. The child had an airway work-up that included laryngoscopy. However, the mass remained undetected. This is most likely explained by the mobile nature of the lesion, prolapsing into the high nasopharynx in supine position. In our patient, magnetic resonance imaging (MRI), initially performed to rule out brainstem pathology, revealed an oropharyngeal dermoid cyst. This case shows the potential role of neuroimaging in the diagnostic work-up of a young child presenting with respiratory distress by excluding a central nervous system lesion and diagnosing an “unexpected” nasopharyngeal mass lesion. In addition, MRI allowed exclusion of skull base lesions of neural origin such as an anterior meningoencephalocele or heterotopic neuroglial tissue which would be managed differently from pharyngeal masses.

Das akute Abdomen ist ein haufiges Krankheitsbild in der Notfalldiagnostik. Pathologische Veranderungen der weiblichen Genitalorgane spielen dabei eine wichtige Rolle. Gynakologische Notfalle gefahrden die Fertilitat und sind potenziell lebensbedrohlich. Bei der Diagnostik sind viele Differenzialdiagnosen je nach Alter der Patientin und einer eventuellen Schwangerschaft zu beachten. Insbesondere akute gastrointestinale Beschwerden lassen sich in der klinischen Untersuchung oft nicht von gynakologischen Notfallen unterscheiden. Die Bildgebung leistet hier einen wesentlichen Beitrag zur Einengung der Differenzialdiagnose, zur Therapieentscheidung und zum Therapiemonitoring. Im vorliegenden Ubersichtsartikel soll der zentrale Stellenwert der Bildgebung im Rahmen haufiger gynakologischer Notfalle erortert werden.

Background: Despite the benefits of whole-body hypothermia therapy, many infants with hypoxic-ischemic encephalopathy (HIE) die or have significant long-term neurodevelopmental impairment. Prospectively identifying neonates at risk of poor outcome is essential but not straightforward. The cerebellum is not classically considered to be a brain region vulnerable to hypoxic-ischemic insults; recent literature suggests, however, that the cerebellum may be involved in neonatal HIE. In this study, we aimed to assess the microstructural integrity of cerebellar and linked supratentorial structures in neonates with HIE compared to neurologically healthy neonatal controls. Methods: In this prospective cohort study, we performed a quantitative diffusion tensor imaging (DTI) analysis of the structural pathways of connectivity, which may be affected in neonatal cerebellar injury by measuring fractional anisotropy (FA) and mean diffusivity (MD) within the superior, middle, and inferior cerebellar peduncles, dentate nuclei, and thalami. All magnetic resonance imaging (MRI) studies were grouped into 4 categories of severity based on a qualitative evaluation of conventional and advanced MRI sequences. Multivariable linear regression analysis of cerebellar scalars of patients and controls was performed, controlling for gestational age, age at the time of MRI, and HIE severity. Spearman rank correlation was performed to correlate DTI scalars of the cerebellum and thalami. Results: Fifty-seven (23 females, 40%) neonates with HIE and 12 (6 females, 50%) neonatal controls were included. There were 8 patients (14%) in HIE severity groups 3 and 4 (injury of the basal ganglia/thalamus and/or cortex). Based on a qualitative analysis of conventional and DTI images, no patients had evidence of cerebellar injury. No significant differences between patients and controls were found in the FA and MD scalars. However, FA values of the middle cerebellar peduncles (0.294 vs. 0.380, p < 0.001) and MD values of the superior cerebellar peduncles (0.920 vs. 1.007 × 10-3 mm/s2, p = 0.001) were significantly lower in patients with evidence of moderate or severe injury on MRI (categories 3 and 4) than in controls. In patients, cerebellar DTI scalars correlated positively with DTI scalars within the thalami. Conclusion: Our results suggest that infants with moderate-to-severe HIE may have occult injury of cerebellar white-matter tracts, which is not detectable by the qualitative analysis of neuroimaging data alone. Cerebellar DTI scalars correlate with thalamic measures, highlighting that cerebellar injury is unlikely to occur in isolation and may reflect the severity of HIE. The impact of concomitant cerebellar injury in HIE on long-term neurodevelopmental outcome warrants further study.

The purpose of this work was to systematically assess the impact of the b-value on texture analysis in MR diffusion-weighted imaging (DWI) of the abdomen. In eight healthy male volunteers, echo-planar DWI sequences at 16 b-values ranging between 0 and 1000 s/mm

Genetic skeletal disorders (GSDs) are a heterogeneous group characterized by an intrinsic abnormality in growth and (re-)modeling of cartilage and bone. A large subgroup of GSDs has additional involvement of other structures/organs beside the skeleton, such as the central nervous system (CNS). CNS abnormalities have an important role in long-term prognosis of children with GSDs and should consequently not be missed. Sensitive and specific identification of CNS lesions while evaluating a child with a GSD requires a detailed knowledge of the possible associated CNS abnormalities. Here, we provide a pattern-recognition approach for neuroimaging findings in GSDs guided by the obvious skeletal manifestations of GSD. In particular, we summarize which CNS findings should be ruled out with each GSD. The diseases (n = 180) are classified based on the skeletal involvement (1. abnormal metaphysis or epiphysis, 2. abnormal size/number of bones, 3. abnormal shape of bones and joints, and 4. abnormal dynamic or structural changes). For each disease, skeletal involvement was defined in accordance with Online Mendelian Inheritance in Man. Morphological CNS involvement has been described based on extensive literature search. Selected examples will be shown based on prevalence of the diseases and significance of the CNS involvement. CNS involvement is common in GSDs. A wide spectrum of morphological abnormalities is associated with GSDs. Early diagnosis of CNS involvement is important in the management of children with GSDs. This pattern-recognition approach aims to assist and guide physicians in the diagnostic work-up of CNS involvement in children with GSDs and their management.