Deep Learning for Brain MRI Segmentation: State of the Art and Future Directions.
Ontology highlight
ABSTRACT:
SUBMITTER: Akkus Z
PROVIDER: S-EPMC5537095 | biostudies-other | 2017 Aug
REPOSITORIES: biostudies-other
Similar Datasets
Project description:Hyperpolarized (HP) carbon 13 (13C) MRI is an emerging molecular imaging method that allows rapid, noninvasive, and pathway-specific investigation of dynamic metabolic and physiologic processes that were previously inaccessible to imaging. This technique has enabled real-time in vivo investigations of metabolism that are central to a variety of diseases, including cancer, cardiovascular disease, and metabolic diseases of the liver and kidney. This review provides an overview of the methods of hyperpolarization and 13C probes investigated to date in preclinical models of disease. The article then discusses the progress that has been made in translating this technology for clinical investigation. In particular, the potential roles and emerging clinical applications of HP [1-13C]pyruvate MRI will be highlighted. The future directions to enable the adoption of this technology to advance the basic understanding of metabolism, to improve disease diagnosis, and to accelerate treatment assessment are also detailed.
| S-EPMC6490043 | biostudies-literature
Project description:Segmentation of brain tissue types from diffusion MRI (dMRI) is an important task, required for quantification of brain microstructure and for improving tractography. Current dMRI segmentation is mostly based on anatomical MRI (e.g., T1- and T2-weighted) segmentation that is registered to the dMRI space. However, such inter-modality registration is challenging due to more image distortions and lower image resolution in dMRI as compared with anatomical MRI. In this study, we present a deep learning method for diffusion MRI segmentation, which we refer to as DDSeg. Our proposed method learns tissue segmentation from high-quality imaging data from the Human Connectome Project (HCP), where registration of anatomical MRI to dMRI is more precise. The method is then able to predict a tissue segmentation directly from new dMRI data, including data collected with different acquisition protocols, without requiring anatomical data and inter-modality registration. We train a convolutional neural network (CNN) to learn a tissue segmentation model using a novel augmented target loss function designed to improve accuracy in regions of tissue boundary. To further improve accuracy, our method adds diffusion kurtosis imaging (DKI) parameters that characterize non-Gaussian water molecule diffusion to the conventional diffusion tensor imaging parameters. The DKI parameters are calculated from the recently proposed mean-kurtosis-curve method that corrects implausible DKI parameter values and provides additional features that discriminate between tissue types. We demonstrate high tissue segmentation accuracy on HCP data, and also when applying the HCP-trained model on dMRI data from other acquisitions with lower resolution and fewer gradient directions.
| S-EPMC8139182 | biostudies-literature
Project description:In the last two decades, neuroscience has produced intriguing evidence for a central role of the claustrum in mammalian forebrain structure and function. However, relatively few in vivo studies of the claustrum exist in humans. A reason for this may be the delicate and sheet-like structure of the claustrum lying between the insular cortex and the putamen, which makes it not amenable to conventional segmentation methods. Recently, Deep Learning (DL) based approaches have been successfully introduced for automated segmentation of complex, subcortical brain structures. In the following, we present a multi-view DL-based approach to segment the claustrum in T1-weighted MRI scans. We trained and evaluated the proposed method in 181 individuals, using bilateral manual claustrum annotations by an expert neuroradiologist as reference standard. Cross-validation experiments yielded median volumetric similarity, robust Hausdorff distance, and Dice score of 93.3%, 1.41 mm, and 71.8%, respectively, representing equal or superior segmentation performance compared to human intra-rater reliability. The leave-one-scanner-out evaluation showed good transferability of the algorithm to images from unseen scanners at slightly inferior performance. Furthermore, we found that DL-based claustrum segmentation benefits from multi-view information and requires a sample size of around 75 MRI scans in the training set. We conclude that the developed algorithm allows for robust automated claustrum segmentation and thus yields considerable potential for facilitating MRI-based research of the human claustrum. The software and models of our method are made publicly available.
| S-EPMC8596988 | biostudies-literature
Project description:Alcohol use disorders remain one of the leading causes of mortality and morbidity across the world, yet despite this impact, there are few treatment options for patients suffering from these disorders. To this end, non-invasive brain stimulation, most commonly utilizing technologies including transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS), has recently emerged as promising potential treatments for alcohol use disorders. Enthusiasm for these interventions is fueled by their non-invasive nature, generally favorable safety profile, and ability to target and modulate brain regions implicated in substance use disorders. In this paper, we describe the underlying principles behind these commonly used stimulation technologies, summarize existing experiments and randomized controlled trials, and provide an integrative summary with suggestions for future areas of research. Currently available data generally supports the use of non-invasive brain stimulation as a near-term treatment for alcohol use disorder, with important caveats regarding the use of stimulation in this patient population.
| S-EPMC7007491 | biostudies-literature
Project description:Gliomas are the most common primary brain malignancies. Accurate and robust tumor segmentation and prediction of patients' overall survival are important for diagnosis, treatment planning and risk factor identification. Here we present a deep learning-based framework for brain tumor segmentation and survival prediction in glioma, using multimodal MRI scans. For tumor segmentation, we use ensembles of three different 3D CNN architectures for robust performance through a majority rule. This approach can effectively reduce model bias and boost performance. For survival prediction, we extract 4,524 radiomic features from segmented tumor regions, then, a decision tree and cross validation are used to select potent features. Finally, a random forest model is trained to predict the overall survival of patients. The 2018 MICCAI Multimodal Brain Tumor Segmentation Challenge (BraTS), ranks our method at 2nd and 5th place out of 60+ participating teams for survival prediction tasks and segmentation tasks respectively, achieving a promising 61.0% accuracy on the classification of short-survivors, mid-survivors and long-survivors.
| S-EPMC6707136 | biostudies-literature
Project description:Background:The growing need for palliative care (PC) among patients with serious illness is outstripped by the short supply of PC specialists. This mismatch calls for competency of all health care providers in primary PC, including patient-centered communication, management of pain and other symptoms, and interprofessional teamwork. Simulation-based medical education (SBME) has emerged as a promising modality to teach key skills and close the educational gap. This paper describes the current state of SBME in training of PC skills. Methods:We conducted a systematic review of the literature reporting on simulation experiences addressing PC skills for clinical learners in medicine and nursing. We collected data on learner characteristics, the method and content of the simulation, and outcome assessments. Results:In a total of 78 studies, 76% involved learners from medicine and 38% involved learners from nursing, while social work (6%) and spiritual care (3%) learners were significantly underrepresented. Only 16% of studies involved collaboration between participants at different training levels. The standardized patient encounter was the most popular simulation method, accounting for 68% of all studies. Eliciting treatment preferences (50%), delivering bad news (41%), and providing empathic communication (40%) were the most commonly addressed skills, while symptom management was only addressed in 13% of studies. The most common method of simulation evaluation was subjective participant feedback (62%). Only 4% of studies examined patient outcomes. In 22% of studies, simulation outcomes were not measured at all. Discussion:We describe the current state of SBME in PC education, highlighting advances over recent decades and identifying gaps and opportunities for future directions. We recommend designing SBME for a broader range of learners and for interprofessional skill building. We advocate for expansion of skill content, especially symptom management education. Finally, evaluation of SBME in PC training should be more rigorous with a shift to include more patient outcomes.
| S-EPMC6292390 | biostudies-literature
Project description:Trauma signature (TSIG) analysis is an evidence-based method that examines the interrelationship between population exposure to a disaster, extreme event, or complex emergency, and the inter-related physical and psychological consequences for the purpose of providing timely, actionable guidance for effective mental health and psychosocial support that is organically tailored and targeted to the defining features of the event. A series of TSIG case studies has been published since 2011 and TSIG analyses of recent disasters are in process. Disaster Health intends to expedite and feature novel TSIG research focusing on late-breaking disaster events. At the current stage of development, expert consensus is sought for refining the TSIG methodology using a Delphi process. The overarching goal is to create a fully operational system to provide timely guidance for adapting disaster behavioral health support to the salient psychological risk factors in each disaster.
| S-EPMC5314909 | biostudies-other
Project description:The purpose of this study was to assess the clinical value of a deep learning (DL) model for automatic detection and segmentation of brain metastases, in which a neural network is trained on four distinct MRI sequences using an input-level dropout layer, thus simulating the scenario of missing MRI sequences by training on the full set and all possible subsets of the input data. This retrospective, multicenter study, evaluated 165 patients with brain metastases. The proposed input-level dropout (ILD) model was trained on multisequence MRI from 100 patients and validated/tested on 10/55 patients, in which the test set was missing one of the four MRI sequences used for training. The segmentation results were compared with the performance of a state-of-the-art DeepLab V3 model. The MR sequences in the training set included pre-gadolinium and post-gadolinium (Gd) T1-weighted 3D fast spin echo, post-Gd T1-weighted inversion recovery (IR) prepped fast spoiled gradient echo, and 3D fluid attenuated inversion recovery (FLAIR), whereas the test set did not include the IR prepped image-series. The ground truth segmentations were established by experienced neuroradiologists. The results were evaluated using precision, recall, Intersection over union (IoU)-score and Dice score, and receiver operating characteristics (ROC) curve statistics, while the Wilcoxon rank sum test was used to compare the performance of the two neural networks. The area under the ROC curve (AUC), averaged across all test cases, was 0.989 ± 0.029 for the ILD-model and 0.989 ± 0.023 for the DeepLab V3 model (p = 0.62). The ILD-model showed a significantly higher Dice score (0.795 ± 0.104 vs. 0.774 ± 0.104, p = 0.017), and IoU-score (0.561 ± 0.225 vs. 0.492 ± 0.186, p < 0.001) compared to the DeepLab V3 model, and a significantly lower average false positive rate of 3.6/patient vs. 7.0/patient (p < 0.001) using a 10 mm3 lesion-size limit. The ILD-model, trained on all possible combinations of four MRI sequences, may facilitate accurate detection and segmentation of brain metastases on a multicenter basis, even when the test cohort is missing input MRI sequences.
| S-EPMC7900111 | biostudies-literature
Project description:IntroductionManagement of patients with brain metastases is often based on manual lesion detection and segmentation by an expert reader. This is a time- and labor-intensive process, and to that end, this work proposes an end-to-end deep learning segmentation network for a varying number of available MRI available sequences.MethodsWe adapt and evaluate a 2.5D and a 3D convolution neural network trained and tested on a retrospective multinational study from two independent centers, in addition, nnU-Net was adapted as a comparative benchmark. Segmentation and detection performance was evaluated by: (1) the dice similarity coefficient, (2) a per-metastases and the average detection sensitivity, and (3) the number of false positives.ResultsThe 2.5D and 3D models achieved similar results, albeit the 2.5D model had better detection rate, whereas the 3D model had fewer false positive predictions, and nnU-Net had fewest false positives, but with the lowest detection rate. On MRI data from center 1, the 2.5D, 3D, and nnU-Net detected 79%, 71%, and 65% of all metastases; had an average per patient sensitivity of 0.88, 0.84, and 0.76; and had on average 6.2, 3.2, and 1.7 false positive predictions per patient, respectively. For center 2, the 2.5D, 3D, and nnU-Net detected 88%, 86%, and 78% of all metastases; had an average per patient sensitivity of 0.92, 0.91, and 0.85; and had on average 1.0, 0.4, and 0.1 false positive predictions per patient, respectively.Discussion/conclusionOur results show that deep learning can yield highly accurate segmentations of brain metastases with few false positives in multinational data, but the accuracy degrades for metastases with an area smaller than 0.4 cm2.
| S-EPMC9889663 | biostudies-literature
Project description:Brain tumor localization and segmentation from magnetic resonance imaging (MRI) are hard and important tasks for several applications in the field of medical analysis. As each brain imaging modality gives unique and key details related to each part of the tumor, many recent approaches used four modalities T1, T1c, T2, and FLAIR. Although many of them obtained a promising segmentation result on the BRATS 2018 dataset, they suffer from a complex structure that needs more time to train and test. So, in this paper, to obtain a flexible and effective brain tumor segmentation system, first, we propose a preprocessing approach to work only on a small part of the image rather than the whole part of the image. This method leads to a decrease in computing time and overcomes the overfitting problems in a Cascade Deep Learning model. In the second step, as we are dealing with a smaller part of brain images in each slice, a simple and efficient Cascade Convolutional Neural Network (C-ConvNet/C-CNN) is proposed. This C-CNN model mines both local and global features in two different routes. Also, to improve the brain tumor segmentation accuracy compared with the state-of-the-art models, a novel Distance-Wise Attention (DWA) mechanism is introduced. The DWA mechanism considers the effect of the center location of the tumor and the brain inside the model. Comprehensive experiments are conducted on the BRATS 2018 dataset and show that the proposed model obtains competitive results: the proposed method achieves a mean whole tumor, enhancing tumor, and tumor core dice scores of 0.9203, 0.9113 and 0.8726 respectively. Other quantitative and qualitative assessments are presented and discussed.
| S-EPMC8149837 | biostudies-literature