Gliomas are the most common types of tumors in the brain. While the definite diagnosis is routinely made ex vivo by histopathologic and molecular examination, diagnostic work-up of patients with suspected glioma is mainly done by using magnetic resonance imaging (MRI). Nevertheless, L-S-methyl-¹¹C-methionine (¹¹C-MET) Positron Emission Tomography (PET) holds a great potential in characterization of gliomas. The aim of this study was to establish machine learning (ML) driven survival models for glioma built on ¹¹C-MET-PET, ex vivo and patient characteristics. Methods: 70 patients with a treatment naïve glioma, who had a positive ¹¹C-MET-PET and histopathology-derived ex vivo feature extraction, such as World Health Organization (WHO) 2007 tumor grade, histology and isocitrate dehydrogenase (IDH1-R132H) mutation status were included. The ¹¹C-MET-positive primary tumors were delineated semi-automatically on PET images followed by the feature extraction of tumor-to-background ratio based general and higher-order textural features by applying five different binning approaches. In vivo and ex vivo features, as well as patient characteristics (age, weight, height, body-mass-index, Karnofsky-score) were merged to characterize the tumors. Machine learning approaches were utilized to identify relevant in vivo, ex vivo and patient features and their relative weights for 36 months survival prediction. The resulting feature weights were used to establish three predictive models per binning configuration based on a combination of: in vivo/ex vivo and clinical patient information (M36IEP), in vivo and patient-only information (M36IP), and in vivo only (M36I). In addition a binning-independent ex vivo and patient-only (M36EP) model was created. The established models were validated in a Monte Carlo (MC) cross-validation scheme. Results: Most prominent ML-selected and -weighted features were patient and ex vivo based followed by in vivo features. The highest area under the curve (AUC) values of our models as revealed by the MC cross-validation were: 0.9 (M36IEP), 0.87 (M36EP), 0.77 (M36IP) and 0.72 (M36I). Conclusion: Survival prediction of glioma patients based on amino acid PET using computer-supported predictive models based on in vivo, ex vivo and patient features is highly accurate.

The software “SyMRI” generates different MR contrasts and characterizes tissue properties based on a single acquisition of a multi-dynamic multi-echo (MDME)-FLAIR sequence. The aim of this study was to assess the applicability of “SyMRI” in the assessment of myelination in preterm and term-born neonates. Furthermore, “SyMRI” was compared with conventional MRI.

Objectives: To characterize spatiotemporal growth differences of prenatal brainstem substructures and cerebellum, using linear biometry and planimetry on fetal magnetic resonance imaging (MRI).Methods: In this retrospective study, we included fetuses with normal brain and a precise midsagittal T2‐weighted brain MRI sequence obtained between May 2003 and April 2019. The cross‐sectional area, rostrocaudal diameter and anteroposterior diameter of the midbrain, pons (basis pontis and pontine tegmentum), medulla oblongata and cerebellar vermis, as well as the transverse cerebellar diameter, were quantified by a single observer. The diameters were also assessed by a second observer to test inter‐rater variability.Results: We included 161 fetuses with normal brain and a precise midsagittal MRI sequence, examined at a mean ± SD gestational age of 25.7 ± 5.4 (range, 14 + 0 to 39 + 2) weeks. All substructures of the fetal brainstem and the cerebellum could be measured consistently (mean ± SD interobserver intraclass correlation coefficient, 0.933 ± 0.065). We provide reference data for diameters and areas of the brainstem and cerebellum in the second and third trimesters. There was a significant quadratic relationship between vermian area and gestational age, and all other measured parameters showed a significant linear growth pattern within the observed period (P < 0.001). A significant change in the relative proportions of the brainstem substructures occurred between the beginning of the second trimester and the end of the third trimester, with an increase in the area of the pons (P < 0.001) and a decrease in that of the midbrain (P < 0.001), relative to the total brainstem area.Conclusions: The substructures of the fetal brainstem follow a distinct spatiotemporal growth pattern, characterized by a relative increase in the pons and decrease in the midbrain, between 15 and 40 weeks of gestation. Caution is needed when interpreting fetal brainstem appearance during the early second trimester, as the brainstem proportions differ significantly from the adult morphology. The reference data provided herein should help to increase diagnostic accuracy in detecting disorders of defective hindbrain segmentation. © 2020 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

Purpose: To investigate the reproducibility of radiomics features extracted from two-dimensional regions of interest (2D ROIs) versus whole lung (3D) ROIs in repeated in-vivo fetal magnetic resonance imaging (MRI) acquisitions. Methods: Thirty fetal MRI scans including two axial T2-weighted acquisitions of the lungs were analysed. 2D (lung at the level of the carina) and 3D (whole lung) ROIs were manually segmented using ITK-Snap. Ninety-five radiomics features were extracted from 2 and 3D ROIs in initial and repeat acquisitions using Pyradiomics. Radiomics feature intra-class correlation coefficients (ICC) were calculated between 2 and 3D ROIs in the initial acquisition, and between 2 and 3D ROIs in repeated acquisitions, respectively. Results: MRI data of 11 (36.7%) female and 19 (63.3%) male fetuses acquired at a median 25 + 0 gestational weeks plus days (GW) (interquartile range [IQR] 23 + 4 − 27 + 0 GW) were assessed. Median radiomics feature ICC between 2 and 3D ROIs in the initial MRI acquisition was 0.733 (IQR 0.313–0.814, range 0.018–0.970). ICCs between radiomics features extracted using 3D ROIs in initial and repeat acquisitions (median 0.908 [IQR 0.824–0.929, range 0.335–0.996]) were significantly higher compared to 2D ROIs (0.771 [0.699–0.835, 0.048–0.965]) (p < 0.001). Conclusion: Fetal MRI radiomics features extracted from 3D whole lung segmentation masks showed significantly higher reproducibility across repeat acquisitions compared to 2D ROIs. Therefore, fetal MRI whole lung radiomics features are robust diagnostic and potentially prognostic tools in the image-based in-vivo quantitative assessment of lung development.

BACKGROUND AND PURPOSE: Temporal lobe epilepsy, structural or nonlesional, may negatively affect language function. However, little is known about the lesion-specific influence on language networks. We hypothesized that different epileptogenic lesions are related to distinct alterations in the functional language connectome detected by fMRI. MATERIALS AND METHODS: One hundred one patients with epilepsy due to mesiotemporal sclerosis (21 left, 22 right), low-grade mesiotemporal tumors (12 left), or nonlesional temporal lobe epilepsy (22 left, 24 right) and 22 healthy subjects performed 3T task-based language fMRI. Task-based activation maps (laterality indices) and functional connectivity analysis (global and connectivity strengths between language areas) were correlated with language scores. RESULTS: Laterality indices based on fMRI activation maps failed to discriminate among patient groups. Functional connectivity analysis revealed the most extended language network alterations in left mesiotemporal sclerosis (involving the left temporal pole, left inferior frontal gyrus, and bilateral premotor areas). The other patient groups showed less extended but also predominantly ipsilesional network changes compared with healthy controls. Left-to-right hippocampal connectivity strength correlated positively with naming function (P = .01), and connectivity strength between the left Wernicke area and the left hippocampus was linked to verbal fluency scores (P = .01) across all groups. CONCLUSIONS: Different pathologies underlying temporal lobe epilepsy are related to distinct alterations of the functional language connectome visualized by fMRI functional connectivity analysis. Network analysis allows new insights into language organization and provides possible imaging biomarkers for language function. These imaging findings emphasize the importance of a personalized treatment strategy in patients with epilepsy.

Purpose: The purpose of this study was to assess the diagnostic accuracy of T1-weighted and T2-weighted contrasts generated by the MR data postprocessing software SyMRI (Synthetic MR AB, Linköping, Sweden) for neonatal brain imaging. Methods: In this study 36 cases of neonatal MRI were retrospectively collected, which included T1-weighted and T2-weighted sequences as well as multi-dynamic multi-echo (MDME) sequences. Of the 36 neonates 32 were included in this study and 4 neuroradiologists independently assessed neonatal brain examinations on the basis of conventional and SyMRI-generated T1-weighted and T2-weighted contrasts, in order to determine the presence or absence of lesions. The sensitivity and specificity of both methods were calculated and compared. Results: Compared to conventionally acquired T1 and T2-weighted images, SyMRI-generated contrasts showed a lower sensitivity but a higher specificity (SyMRI sensitivity 0.88, confidence interval (CI): 0.72–0.95; specificity 1, CI: 0.89–1/conventional MRI: sensitivity: 0.94, CI: 0.80–0.98; specificity: 0.94, CI: 0.80–0.98). Conclusion: The T1-weighted and T2-weighted images generated by SyMRI showed a diagnostic accuracy comparable to that of conventionally acquired contrasts. In addition to semiquantitative imaging data, SyMRI provides diagnostic images and leads to a more efficient use of available imaging time in neonatal brain MRI.