Search

Brain segmentation from neonatal MRI images is a very challenging task due to large changes in the shape of cerebral structures and variations in signal intensities reflecting the gestational process. In this context, there is a clear need for segmentation techniques that are robust to variations in image contrast and to the spatial configuration of anatomical structures. In this work, we evaluate the potential of synthetic learning, a contrast‐independent model trained using synthetic images generated from the ground truth labels of very few subjects. We base our experiments on the dataset released by the developmental Human Connectome Project, for which high‐quality images are available for more than 700 babies aged between 26 and 45 weeks postconception. First, we confirm the impressive performance of a standard UNet trained on a few volumes, but also confirm that such models learn intensity‐related features specific to the training domain. We then confirm the robustness of the synthetic learning approach to variations in image contrast. However, we observe a clear influence of the age of the baby on the predictions. We improve the performance of this model by enriching the synthetic training set with realistic motion artifacts and over‐segmentation of the white matter. Based on extensive visual assessment, we argue that the better performance of the model trained on real T2w data may be due to systematic errors in the ground truth. We propose an original experiment allowing us to show that learning from real data will reproduce any systematic bias affecting the training set, while synthetic models can avoid this limitation. Overall, our experiments confirm that synthetic learning is an effective solution for segmenting neonatal brain MRI. Our adapted synthetic learning approach combines key features that will be instrumental for large multisite studies and clinical applications. [ABSTRACT FROM AUTHOR]

Background: Accurate segmentation of brain glioma is a critical prerequisite for clinical diagnosis, surgical planning and treatment evaluation. In current clinical workflow, physicians typically perform delineation of brain tumor subregions slice‐by‐slice, which is more susceptible to variabilities in raters and also time‐consuming. Besides, even though convolutional neural networks (CNNs) are driving progress, the performance of standard models still have some room for further improvement. Purpose: To deal with these issues, this paper proposes an attention‐guided multi‐scale context aggregation network (AMCA‐Net) for the accurate segmentation of brain glioma in the magnetic resonance imaging (MRI) images with multi‐modalities. Methods: AMCA‐Net extracts the multi‐scale features from the MRI images and fuses the extracted discriminative features via a self‐attention mechanism for brain glioma segmentation. The extraction is performed via a series of down‐sampling, convolution layers, and the global context information guidance (GCIG) modules are developed to fuse the features extracted for contextual features. At the end of the down‐sampling, a multi‐scale fusion (MSF) module is designed to exploit and combine all the extracted multi‐scale features. Each of the GCIG and MSF modules contain a channel attention (CA) module that can adaptively calibrate feature responses and emphasize the most relevant features. Finally, multiple predictions with different resolutions are fused through different weightings given by a multi‐resolution adaptation (MRA) module instead of the use of averaging or max‐pooling to improve the final segmentation results. Results: Datasets used in this paper are publicly accessible, that is, the Multimodal Brain Tumor Segmentation Challenges 2018 (BraTS2018) and 2019 (BraTS2019). BraTS2018 contains 285 patient cases and BraTS2019 contains 335 cases. Simulations show that the AMCA‐Net has better or comparable performance against that of the other state‐of‐the‐art models. In terms of the Dice score and Hausdorff 95 for the BraTS2018 dataset, 90.4% and 10.2 mm for the whole tumor region (WT), 83.9% and 7.4 mm for the tumor core region (TC), 80.2% and 4.3 mm for the enhancing tumor region (ET), whereas the Dice score and Hausdorff 95 for the BraTS2019 dataset, 91.0% and 10.7 mm for the WT, 84.2% and 8.4 mm for the TC, 80.1% and 4.8 mm for the ET. Conclusions: The proposed AMCA‐Net performs comparably well in comparison to several state‐of‐the‐art neural net models in identifying the areas involving the peritumoral edema, enhancing tumor, and necrotic and non‐enhancing tumor core of brain glioma, which has great potential for clinical practice. In future research, we will further explore the feasibility of applying AMCA‐Net to other similar segmentation tasks. [ABSTRACT FROM AUTHOR]

Background: Fast multi‐contrast echo planar MRI (EPIMix) has comparable diagnostic performance to standard MRI for detecting brain pathology but its performance in detecting acute cerebral infarctions has not been determined. Purpose: To assess the diagnostic performance of EPIMix for the detection of acute cerebral infarctions. Study Type: Retrospective observational cohort. Population: One hundred and seventy‐two consecutive patients with a clinical suspicion of non‐hyperacute ischemic stroke (January 2018 to December 2019). Field Strength and Sequence: 1.5 T or 3 T. EPIMix ((echo‐planar based: diffusion weighted (DWI), T2*‐weighted, T2‐weighted, T2‐ and T1‐fluid attenuated inversion recovery (FLAIR) images) vs. standard MRI: echo‐planar DWI, echo‐planar T2*‐weighted or susceptibility weighted, turbo spin‐echo T2‐weighted, T2‐ and T1‐FLAIR turbo spin‐echo sequences. Assessment: Three neuroradiologists rated EPIMix and standard MRI on two separate occasions. Incongruent assessments were resolved in consensus with the fourth reader. The ratings included the diagnostic category (acute infarct, normal, and other pathology). Congruent diagnoses together with consensus diagnoses served as the reference standard. Statistical Tests: The diagnostic performance of EPIMix and standard MRI against the reference standard was calculated by the area under the receiver operating characteristic curve (AUC) and compared by DeLong's test. Sensitivity and specificity were determined. Inter‐rater agreements were evaluated by Fleiss's kappa. Results: Of 172 patients (61 ± 16 years, 103 men), acute infarcts were present in 80/172 (47%), normal findings in 60/172 (35%), and other pathology in 32/172 (19%). Across readers, the AUCs were.94–.95 for EPIMix and.95–.99 for standard MRI, with overlapping 95% CI (P =.02–.18). Inter‐rater agreement for EPIMix was 0.90 and for standard MRI was 0.93. The sensitivity for EPIMix and standard MRI was 88–91% and 91–98%, respectively, while the specificity was 98–100% and 98–99%, both with overlapping 95% CI. Conclusion: Multi‐contrast echo planar MRI showed a high but marginally lower diagnostic performance compared to standard MRI for the detection and characterization of acute brain infarct. Level of Evidence: 3 Technical Efficacy: Stage 2 [ABSTRACT FROM AUTHOR]