Brain Lesion Segmentation Research Articles

CarveMix: A simple data augmentation method for brain lesion segmentation

Brain lesion segmentation provides a valuable tool for clinical diagnosis and research, and convolutional neural networks (CNNs) have achieved unprecedented success in the segmentation task. Data augmentation is a widely used strategy to improve the training of CNNs. In particular, data augmentation approaches that mix pairs of annotated training images have been developed. These methods are easy to implement and have achieved promising results in various image processing tasks. However, existing data augmentation approaches based on image mixing are not designed for brain lesions and may not perform well for brain lesion segmentation. Thus, the design of this type of simple data augmentation method for brain lesion segmentation is still an open problem. In this work, we propose a simple yet effective data augmentation approach, dubbed as CarveMix, for CNN-based brain lesion segmentation. Like other mixing-based methods, CarveMix stochastically combines two existing annotated images (annotated for brain lesions only) to obtain new labeled samples. To make our method more suitable for brain lesion segmentation, CarveMix is lesion-aware, where the image combination is performed with a focus on the lesions and preserves the lesion information. Specifically, from one annotated image we carve a region of interest (ROI) according to the lesion location and geometry with a variable ROI size. The carved ROI then replaces the corresponding voxels in a second annotated image to synthesize new labeled images for network training, and additional harmonization steps are applied for heterogeneous data where the two annotated images can originate from different sources. Besides, we further propose to model the mass effect that is unique to whole brain tumor segmentation during image mixing. To evaluate the proposed method, experiments were performed on multiple publicly available or private datasets, and the results show that our method improves the accuracy of brain lesion segmentation. The code of the proposed method is available at https://github.com/ZhangxinruBIT/CarveMix.git.

Two stages cascades neural network for multi-class brain lesion classification system in MRI images

Brain tumor is one of the deadliest cancerous diseases and their severity has turned them into the leading cause of cancer-related mortality. Automatic detection and classification of severity-level for a brain tumor using MRI is a complex process in multilevel classification and needs an improved learning method without computational complexity. In this research article, we propose an innovative Multi-Dimensional Cascades Neural Network work (MDCNet) that takes full advantage of two networks with different dimensions, which can balance the complete semantic information and high-resolution detail information of a large-volume MRI image. In stage 1, a shallow-layer-enhanced 3D location net obtains the location and rough segmentation of brain lesions. In stage 2, a high-resolution attention map is used to obtain the 2D high-resolution image slice sets from the original image and the output of stage 1. The high-resolution images pick up the lost detailed information, refining the boundaries further. Moreover, a multi-view 2.5D net composed of three 2D refinement sub-networks is applied to deeply explore the morphological characteristics of all brain lesions from different perspectives, which compensates for the mistakes and missing spatial information of a single view, increasing the stability of the whole algorithm. The robustness of the proposed model is analyzed using several performance metrics of three different data sets. Through the prominent performance, the proposed model can outperform other existing models attaining an average accuracy of 99.13%. Here, the individual accuracy for Dataset 1, Dataset 2, and Dataset 3 is 99.67%, 98.16%, and 99.76% respectively.

Label fusion and training methods for reliable representation of inter-rater uncertainty

Medical tasks are prone to inter-rater variability due to multiple factors such as image quality, professional experience and training, or guideline clarity. Training deep learning networks with annotations from multiple raters is a common practice that mitigates the model’s bias towards a single expert. Reliable models generating calibrated outputs and reflecting the inter-rater disagreement are key to the integration of artificial intelligence in clinical practice. Various methods exist to take into account different expert labels. We focus on comparing three label fusion methods: STAPLE, average of the rater’s segmentation, and random sampling of each rater’s segmentation during training. Each label fusion method is studied using both the conventional training framework and the recently published SoftSeg framework that limits information loss by treating the segmentation task as a regression. Our results, across 10 data splittings on two public datasets (spinal cord gray matter challenge, and multiple sclerosis brain lesion segmentation), indicate that SoftSeg models, regardless of the ground truth fusion method, had better calibration and preservation of the inter-rater rater variability compared with their conventional counterparts without impacting the segmentation performance. Conventional models, i.e., trained with a Dice loss, with binary inputs, and sigmoid/softmax final activate, were overconfident and underestimated the uncertainty associated with inter-rater variability. Conversely, fusing labels by averaging with the SoftSeg framework led to underconfident outputs and overestimation of the rater disagreement. In terms of segmentation performance, the best label fusion method was different for the two datasets studied, indicating this parameter might be task-dependent. However, SoftSeg had segmentation performance systematically superior or equal to the conventionally trained models and had the best calibration and preservation of the inter-rater variability. SoftSeg has a low computational cost and performed similarly in terms of uncertainty to ensembles which require multiple models and forward passes. Our code is available at <href='https://ivadomed.org'>https://ivadomed.org</a>.

Automatic localisation and per-region quantification of traumatic brain injury on head CT using atlas mapping

Rationale and objectivesTo develop a method for automatic localisation of brain lesions on head CT, suitable for both population-level analysis and lesion management in a clinical setting. Materials and methodsLesions were located by mapping a bespoke CT brain atlas to the patient’s head CT in which lesions had been previously segmented. The atlas mapping was achieved through robust intensity-based registration enabling the calculation of per-region lesion volumes. Quality control (QC) metrics were derived for automatic detection of failure cases. The CT brain template was built using 182 non-lesioned CT scans and an iterative template construction strategy. Individual brain regions in the CT template were defined via non-linear registration of an existing MRI-based brain atlas.Evaluation was performed on a multi-centre traumatic brain injury dataset (TBI) (n = 839 scans), including visual inspection by a trained expert. Two population-level analyses are presented as proof-of-concept: a spatial assessment of lesion prevalence, and an exploration of the distribution of lesion volume per brain region, stratified by clinical outcome. Results95.7% of the lesion localisation results were rated by a trained expert as suitable for approximate anatomical correspondence between lesions and brain regions, and 72.5% for more quantitatively accurate estimates of regional lesion load. The classification performance of the automatic QC showed an AUC of 0.84 when compared to binarised visual inspection scores. The localisation method has been integrated into the publicly available Brain Lesion Analysis and Segmentation Tool for CT (BLAST-CT). ConclusionAutomatic lesion localisation with reliable QC metrics is feasible and can be used for patient-level quantitative analysis of TBI, as well as for large-scale population analysis due to its computational efficiency (<2 min/scan on GPU).

A two-stage U-net approach to brain tumor segmentation from multi-spectral MRI records

Abstract The automated segmentation of brain tissues and lesions represents a widely investigated research topic. The Brain Tumor Segmentation Challenges (BraTS) organized yearly since 2012 provided standard training and testing data and a unified evaluation framework to the research community, which provoked an intensification in this research field. This paper proposes a solution to the brain tumor segmentation problem, which is built upon the U-net architecture that is very popular in medical imaging. The proposed procedure involves two identical, cascaded U-net networks with 3D convolution. The first stage produces an initial segmentation of a brain volume, while the second stage applies a post-processing based on the labels provided by the first stage. In the first U-net based classification, each pixel is characterized by the four observed features (T1, T2, T1c, and FLAIR), while the second identical U-net works with four features extracted from the volumetric neighborhood of the pixels, representing the ratio of pixels with positive initial labeling within the neighborhood. Statistical accuracy indexes are employed to evaluate the initial and final segmentation of each MRI record. Tests based on BraTS 2019 training data set led to average Dice scores over 87%. The postprocessing step can increase the average Dice scores by 0.5%, it improves more those volumes whose initial segmentation was less successful.

Semantic Segmentation of Spontaneous Intracerebral Hemorrhage, Intraventricular Hemorrhage, and Associated Edema on CT Images Using Deep Learning.

This study evaluated deep learning algorithms for semantic segmentation and quantification of intracerebral hemorrhage (ICH), perihematomal edema (PHE), and intraventricular hemorrhage (IVH) on noncontrast CT scans of patients with spontaneous ICH. Models were assessed on 1732 annotated baseline noncontrast CT scans obtained from the Tranexamic Acid for Hyperacute Primary Intracerebral Haemorrhage (ie, TICH-2) international multicenter trial (ISRCTN93732214), and different loss functions using a three-dimensional no-new-U-Net (nnU-Net) were examined to address class imbalance (30% of participants with IVH in dataset). On the test cohort (n = 174, 10% of dataset), the top-performing models achieved median Dice similarity coefficients of 0.92 (IQR, 0.89-0.94), 0.66 (0.58-0.71), and 1.00 (0.87-1.00), respectively, for ICH, PHE, and IVH segmentation. U-Net-based networks showed comparable, satisfactory performances on ICH and PHE segmentations (P > .05), but all nnU-Net variants achieved higher accuracy than the Brain Lesion Analysis and Segmentation Tool for CT (BLAST-CT) and DeepLabv3+ for all labels (P < .05). The Focal model showed improved performance in IVH segmentation compared with the Tversky, two-dimensional nnU-Net, U-Net, BLAST-CT, and DeepLabv3+ models (P < .05). Focal achieved concordance values of 0.98, 0.88, and 0.99 for ICH, PHE, and ICH volumes, respectively. The mean volumetric differences between the ground truth and prediction were 0.32 mL (95% CI: -8.35, 9.00), 1.14 mL (-9.53, 11.8), and 0.06 mL (-1.71, 1.84), respectively. In conclusion, U-Net-based networks provide accurate segmentation on CT images of spontaneous ICH, and Focal loss can address class imbalance. International Clinical Trials Registry Platform (ICTRP) no. ISRCTN93732214 Supplemental material is available for this article. © RSNA, 2022 Keywords: Head/Neck, Brain/Brain Stem, Hemorrhage, Segmentation, Quantification, Convolutional Neural Network (CNN), Deep Learning Algorithms, Machine Learning Algorithms.

Constrained unsupervised anomaly segmentation.

Current unsupervised anomaly localization approaches rely on generative models to learn the distribution of normal images, which is later used to identify potential anomalous regions derived from errors on the reconstructed images. To address the limitations of residual-based anomaly localization, very recent literature has focused on attention maps, by integrating supervision on them in the form of homogenization constraints. In this work, we propose a novel formulation that addresses the problem in a more principled manner, leveraging well-known knowledge in constrained optimization. In particular, the equality constraint on the attention maps in prior work is replaced by an inequality constraint, which allows more flexibility. In addition, to address the limitations of penalty-based functions we employ an extension of the popular log-barrier methods to handle the constraint. Last, we propose an alternative regularization term that maximizes the Shannon entropy of the attention maps, reducing the amount of hyperparameters of the proposed model. Comprehensive experiments on two publicly available datasets on brain lesion segmentation demonstrate that the proposed approach substantially outperforms relevant literature, establishing new state-of-the-art results for unsupervised lesion segmentation.

Optimized activation for quantum-inspired self-supervised neural network based fully automated brain lesion segmentation

Optimized activation for quantum-inspired self-supervised neural network based fully automated brain lesion segmentation

Segmentation of Human Brain Gliomas Tumour Images using U-Net Architecture with Transfer Learning

The complexity of segmenting a brain tumour is critical in medical image processing. Treatment options and patient survival rates can only be improved if brain tumours can be prevented and treated. Segmentation of the brain is the most complex and time-consuming task to diagnose cancer utilizing a manual approach for numerous magnetic resonance images (MRI). The aim of MRI brain tumour image segmentation that to build an automated magnetic resonance imaging tumour segmentation system with separate the area of tumour and provided a clear boundary of the tumour region. U-Nets with different transfer learning models as backbones are presented in this paper, there are ResNet50, DenseNet169 and EfficientNet-B7. Brain lesion segmentation is performed using the multimodal brain tumor segmentation challenge 2020 dataset (BraTS2020). Based on MRI scans of the brain, the tumor segmentation technique is assessed using F1-score, Dice loss, and intersection over union score (IoU). The U-Net encoder used with EfficientNet-B7 outperforms all other architectures in terms of performance metrics across the board. Overall, the results of this experiment are rather excellent. The Dice-loss score was 0.009435, and the score of IoU was 0.7435, F1-score was 0.9848, accuracy was 0.9924, precision was 0.9829, recall was 0.9868, and specificity was 0.9943. The U-Net with EfficientNet-B7 architecture was shown to be crucial in the treatment of brain tumors, according to the findings of the experiments

Structural asymmetries in normal brain anatomy: A brief overview

PurposeA fundamental feature in interpreting gross or neuroimaging brain anatomy is reliance on an assumed high degree of morphologic symmetry in bilateral hemispheres. However, the normal brain is not perfectly symmetrical, and subtle inherent structural asymmetries could potentially confound appreciation of pathology-induced asymmetry or how a given brain asymmetry can relate to its function. Material and MethodsWe review the literature and provide a brief overview of structural asymmetries in normal brain anatomy. ResultsBrain structural asymmetries are either rotational or pure right-left asymmetries, and many are a consequence of unique features linked to the use of human language. Yakovlevian torque is the tendency of the right hemisphere to rotate slightly forward relative to the left, which may make the right frontal lobe bigger and wider, and the left occipital lobe wider and protrude rightward. This makes the left Sylvian fissure longer and flatter, resulting in a larger planum temporale. We also discuss right-left asymmetries in the cortex, white matter structures, deep gray nuclei, and lateral ventricles. Brain asymmetries are not random but result from distinct patterns in structural design that confer evolutionary functional advantages. ConclusionMinor brain asymmetries are important and should be accounted for as they can be connected to function, and like individual variability, are essential for evolution. This overview will help understand structural brain asymmetries for improved diagnostic neuroimaging interpretation, constructing symmetry-based paradigms for automatic localization, segmentation of brain lesions, and as a reference for studies on possible implications of excessive asymmetry and altered laterality in cognitive, neurological, and psychiatric disorders.

Lesion Volume Quantification Using Two Convolutional Neural Networks in MRIs of Multiple Sclerosis Patients.

Background: Multiple sclerosis (MS) is a neurologic disease of the central nervous system which affects almost three million people worldwide. MS is characterized by a demyelination process that leads to brain lesions, allowing these affected areas to be visualized with magnetic resonance imaging (MRI). Deep learning techniques, especially computational algorithms based on convolutional neural networks (CNNs), have become a frequently used algorithm that performs feature self-learning and enables segmentation of structures in the image useful for quantitative analysis of MRIs, including quantitative analysis of MS. To obtain quantitative information about lesion volume, it is important to perform proper image preprocessing and accurate segmentation. Therefore, we propose a method for volumetric quantification of lesions on MRIs of MS patients using automatic segmentation of the brain and lesions by two CNNs. Methods: We used CNNs at two different moments: the first to perform brain extraction, and the second for lesion segmentation. This study includes four independent MRI datasets: one for training the brain segmentation models, two for training the lesion segmentation model, and one for testing. Results: The proposed brain detection architecture using binary cross-entropy as the loss function achieved a 0.9786 Dice coefficient, 0.9969 accuracy, 0.9851 precision, 0.9851 sensitivity, and 0.9985 specificity. In the second proposed framework for brain lesion segmentation, we obtained a 0.8893 Dice coefficient, 0.9996 accuracy, 0.9376 precision, 0.8609 sensitivity, and 0.9999 specificity. After quantifying the lesion volume of all patients from the test group using our proposed method, we obtained a mean value of 17,582 mm3. Conclusions: We concluded that the proposed algorithm achieved accurate lesion detection and segmentation with reproducibility corresponding to state-of-the-art software tools and manual segmentation. We believe that this quantification method can add value to treatment monitoring and routine clinical evaluation of MS patients.

Energy functional driven by multiple features for brain lesion segmentation

Energy functional driven by multiple features for brain lesion segmentation

Visualizing MRI Deep Learning Segmentation Algorithms using 3D Printing

As the capabilities and roles of Artificial Intelligence (AI) in the medical field are continually expanded, new potential uses, and combinations of technology become viable. This paper highlights a methodology for utilizing AI Magnetic Resonance Imaging (MRI) segmentation networks and 3D printing processes in conjunction for medical diagnosis, planning, and visualization of medical images. We also include promising benefits and potential medical offerings made possible by this system. By training a "U- Net" on the 2019 BraTS dataset, we base our research on an MRI brain lesion segmentation dataset with sustaining performance and world recognition. This network automatically segments novel MRI scans into lesion and non-lesion regions. We pair this network with a 3D printing process that enables us to print fully segmented, 1:1 scale, patient organs aided by AI techniques to better explain cases, test models, and plan for operations. In addition to a clearly outlined process that enables these offerings, we establish a potential trajectory for how these combined tools will continue to revolutionize the ways healthcare professionals interact with patients and their data.

Deep Learning-Based Segmentation of Various Brain Lesions for Radiosurgery

Semantic segmentation of medical images with deep learning models is rapidly being developed. In this study, we benchmarked state-of-the-art deep learning segmentation algorithms on our clinical stereotactic radiosurgery dataset. The dataset consists of 1688 patients with various brain lesions (pituitary tumors, meningioma, schwannoma, brain metastases, arteriovenous malformation, and trigeminal neuralgia), and we divided the dataset into a training set (1557 patients) and test set (131 patients). This study demonstrates the strengths and weaknesses of deep-learning algorithms in a fairly practical scenario. We compared the model performances concerning their sampling method, model architecture, and the choice of loss functions, identifying suitable settings for their applications and shedding light on the possible improvements. Evidence from this study led us to conclude that deep learning could be promising in assisting the segmentation of brain lesions even if the training dataset was of high heterogeneity in lesion types and sizes.

MTANS: Multi-Scale Mean Teacher Combined Adversarial Network with Shape-Aware Embedding for Semi-Supervised Brain Lesion Segmentation

The annotation of brain lesion images is a key step in clinical diagnosis and treatment of a wide spectrum of brain diseases. In recent years, segmentation methods based on deep learning have gained unprecedented popularity, leveraging a large amount of data with high-quality voxel-level annotations. However, due to the limited time clinicians can provide for the cumbersome task of manual image segmentation, semi-supervised medical image segmentation methods present an alternative solution as they require only a few labeled samples for training. In this paper, we propose a novel semi-supervised segmentation framework that combines improved mean teacher and adversarial network. Specifically, our framework consists of (i) a student model and a teacher model for segmenting the target and generating the signed distance maps of object surfaces, and (ii) a discriminator network for extracting hierarchical features and distinguishing the signed distance maps of labeled and unlabeled data. Besides, based on two different adversarial learning processes, a multi-scale feature consistency loss derived from the student and teacher models is proposed, and a shape-aware embedding scheme is integrated into our framework. We evaluated the proposed method on the public brain lesion datasets from ISBI 2015, ISLES 2015, and BRATS 2018 for the multiple sclerosis lesion, ischemic stroke lesion, and brain tumor segmentation respectively. Experiments demonstrate that our method can effectively leverage unlabeled data while outperforming the supervised baseline and other state-of-the-art semi-supervised methods trained with the same labeled data. The proposed framework is suitable for joint training of limited labeled data and additional unlabeled data, which is expected to reduce the effort of obtaining annotated images.

Multiple Sclerosis Lesion Segmentation in Brain MRI Using Inception Modules Embedded in a Convolutional Neural Network.

Multiple sclerosis (MS) is a chronic and autoimmune disease that forms lesions in the central nervous system. Quantitative analysis of these lesions has proved to be very useful in clinical trials for therapies and assessing disease prognosis. However, the efficacy of these quantitative analyses greatly depends on how accurately the MS lesions have been identified and segmented in brain MRI. This is usually carried out by radiologists who label 3D MR images slice by slice using commonly available segmentation tools. However, such manual practices are time consuming and error prone. To circumvent this problem, several automatic segmentation techniques have been investigated in recent years. In this paper, we propose a new framework for automatic brain lesion segmentation that employs a novel convolutional neural network (CNN) architecture. In order to segment lesions of different sizes, we have to pick a specific filter or size 3 × 3 or 5 × 5. Sometimes, it is hard to decide which filter will work better to get the best results. Google Net has solved this problem by introducing an inception module. An inception module uses 3 × 3, 5 × 5, 1 × 1 and max pooling filters in parallel fashion. Results show that incorporating inception modules in a CNN has improved the performance of the network in the segmentation of MS lesions. We compared the results of the proposed CNN architecture for two loss functions: binary cross entropy (BCE) and structural similarity index measure (SSIM) using the publicly available ISBI-2015 challenge dataset. A score of 93.81 which is higher than the human rater with BCE loss function is achieved.

The Effect of Slice Thickness on Contours of Brain Metastases for Stereotactic Radiosurgery

ObjectivesStereotactic radiosurgery is a common treatment for brain metastases and is typically planned on magnetic resonance imaging (MRI). However, the MR acquisition parameters used for patient selection and treatment planning for stereotactic radiosurgery can vary within and across institutions. In this work, we investigate the effect of MRI slice thickness on the detection and contoured volume of metastatic lesions in the brain. Methods and MaterialsA retrospective cohort of 28 images acquired with a slice thickness of 1 mm were resampled to simulate acquisitions at 2- and 3-mm slice thickness. A total of 102 metastases ranging from 0.0030 cc to 5.08 cc (75-percentile 0.36 cc) were contoured on the original images. All 3 sets of images were recontoured by experienced physicians. ResultsOf all the images detected and contoured on the 1 mm images, 3% of lesions were missed on the 2 mm images, and 13% were missed on the 3 mm images. One lesion that was identified on both the 2 mm and 3 mm images was determined to be a blood vessel on the 1 mm images. Additionally, the lesions were contoured 11% larger on the 2 mm and 43% larger on the 3 mm images. ConclusionsUsing images with a slice thickness >1 mm effects detection and segmentation of brain lesions, which can have an important effect on patient management and treatment outcomes.

Generative Adversarial Networks to Synthesize Missing T1 and FLAIR MRI Sequences for Use in a Multisequence Brain Tumor Segmentation Model.

Background Missing MRI sequences represent an obstacle in the development and use of deep learning (DL) models that require multiple inputs. Purpose To determine if synthesizing brain MRI scans using generative adversarial networks (GANs) allows for the use of a DL model for brain lesion segmentation that requires T1-weighted images, postcontrast T1-weighted images, fluid-attenuated inversion recovery (FLAIR) images, and T2-weighted images. Materials and Methods In this retrospective study, brain MRI scans obtained between 2011 and 2019 were collected, and scenarios were simulated in which the T1-weighted images and FLAIR images were missing. Two GANs were trained, validated, and tested using 210 glioblastomas (GBMs) (Multimodal Brain Tumor Image Segmentation Benchmark [BRATS] 2017) to generate T1-weighted images from postcontrast T1-weighted images and FLAIR images from T2-weighted images. The quality of the generated images was evaluated with mean squared error (MSE) and the structural similarity index (SSI). The segmentations obtained with the generated scans were compared with those obtained with the original MRI scans using the dice similarity coefficient (DSC). The GANs were validated on sets of GBMs and central nervous system lymphomas from the authors' institution to assess their generalizability. Statistical analysis was performed using the Mann-Whitney, Friedman, and Dunn tests. Results Two hundred ten GBMs from the BRATS data set and 46 GBMs (mean patient age, 58 years ± 11 [standard deviation]; 27 men [59%] and 19 women [41%]) and 21 central nervous system lymphomas (mean patient age, 67 years ± 13; 12 men [57%] and nine women [43%]) from the authors' institution were evaluated. The median MSE for the generated T1-weighted images ranged from 0.005 to 0.013, and the median MSE for the generated FLAIR images ranged from 0.004 to 0.103. The median SSI ranged from 0.82 to 0.92 for the generated T1-weighted images and from 0.76 to 0.92 for the generated FLAIR images. The median DSCs for the segmentation of the whole lesion, the FLAIR hyperintensities, and the contrast-enhanced areas using the generated scans were 0.82, 0.71, and 0.92, respectively, when replacing both T1-weighted and FLAIR images; 0.84, 0.74, and 0.97 when replacing only the FLAIR images; and 0.97, 0.95, and 0.92 when replacing only the T1-weighted images. Conclusion Brain MRI scans generated using generative adversarial networks can be used as deep learning model inputs in case MRI sequences are missing. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Zhong in this issue. An earlier incorrect version of this article appeared online. This article was corrected on April 12, 2021.

An improved framework for brain tumor analysis using MRI based on YOLOv2 and convolutional neural network

Brain tumor is a group of anomalous cells. The brain is enclosed in a more rigid skull. The abnormal cell grows and initiates a tumor. Detection of tumor is a complicated task due to irregular tumor shape. The proposed technique contains four phases, which are lesion enhancement, feature extraction and selection for classification, localization, and segmentation. The magnetic resonance imaging (MRI) images are noisy due to certain factors, such as image acquisition, and fluctuation in magnetic field coil. Therefore, a homomorphic wavelet filer is used for noise reduction. Later, extracted features from inceptionv3 pre-trained model and informative features are selected using a non-dominated sorted genetic algorithm (NSGA). The optimized features are forwarded for classification after which tumor slices are passed to YOLOv2-inceptionv3 model designed for the localization of tumor region such that features are extracted from depth-concatenation (mixed-4) layer of inceptionv3 model and supplied to YOLOv2. The localized images are passed to McCulloch's Kapur entropy method to segment actual tumor region. Finally, the proposed technique is validated on three benchmark databases BRATS 2018, BRATS 2019, and BRATS 2020 for tumor detection. The proposed method achieved greater than 0.90 prediction scores in localization, segmentation and classification of brain lesions. Moreover, classification and segmentation outcomes are superior as compared to existing methods.

Abstract 10: Radiomic Signature of the White Matter Hyperintensity Burden Correlates With Clinical Phenotypes

Introduction: Structural integrity of cerebral parenchyma is an essential radiographic equivalent of brain health; but its assessment usually requires dedicated advanced image acquisitions. Radiomics analyses bear the potential to describe radiophenotypes beyond what meets the naked eye. We sought to: 1) evaluate this novel approach to predict white matter hyperintensity (WMH) burden and 2) uncover latent clinico-radiological associations. Methods: An international, multi-site cohort of 4,164 acute ischemic stroke (AIS) patients with FLAIR MRI (MRI-GENIE study) underwent total brain and WMH lesion segmentation using convolutional neural networks. Radiomic features (n=1905) were extracted from clinical FLAIR images outside of the WMH (brain mask - WMH mask). Prediction of the WMH burden using radiomics was done using LASSO regression. Radiomic signature of WMH was built with the most stable selected features, then compared to the clinical variables using canonical correlation analysis. Results: In this cohort, (mean age=62.8±15.0, median WMH volume=4.2cc IQR 1.4-11.2), radiomic features were highly predictive of WMH burden (R2=0.8±0.012). Radiomic signature of WMH included 68 features. All 7 pairs of extracted canonical variates (CV) were statistically significant with respective canonical correlations of 0.79, 0.64, 0.44, 0.21, 0.16, 0.15 (Bonferroni corrected p-values CV1-6 &lt;.001, p-value CV7 =.003). Upon examination, CV1 was mainly influenced by age, CV2 by sex, CV3 by history of smoking and diabetes mellitus (DM), CV4 by hypertension, CV5 by atrial fibrillation (AF) and DM, CV6 by coronary artery disease (CAD) and CV7 by CAD and DM. Conclusion: Radiomics extracted from clinical grade FLAIR images of AIS patients seem able to capture structural integrity of the cerebral parenchyma and to correlate with clinical phenotypes. Further research could evaluate radiomics to predict the progression of cerebral small vessel disease on longitudinal data.