Skull-Stripping of Glioblastoma MRI Scans Using 3D Deep Learning

Accuracy of skull stripping in a single-contrast convolutional neural network model using eight-contrast magnetic resonance images

In automated analyses of brain morphometry, skull stripping or brain extraction is a critical first step because it provides accurate spatial registration and signal-intensity normalization. Therefore, it is imperative to develop an ideal skull-stripping method in the field of brain image analysis. Previous reports have shown that convolutional neural network (CNN) method is better at skull stripping than non-CNN methods. We aimed to evaluate the accuracy of skull stripping in a single-contrast CNN model using eight-contrast magnetic resonance (MR) images. A total of 12 healthy participants and 12 patients with a clinical diagnosis of unilateral Sturge-Weber syndrome were included in our study. A 3-T MR imaging system and QRAPMASTER were used for data acquisition. We obtained eight-contrast images produced by post-processing T1, T2, and proton density (PD) maps. To evaluate the accuracy of skull stripping in our CNN method, gold-standard intracranial volume (ICVG) masks were used to train the CNN model. The ICVG masks were defined by experts using manual tracing. The accuracy of the intracranial volume obtained from the single-contrast CNN model (ICVE) was evaluated using the Dice similarity coefficient [= 2(ICVE ⋂ ICVG)/(ICVE + ICVG)]. Our study showed significantly higher accuracy in the PD-weighted image (WI), phase-sensitive inversion recovery (PSIR), and PD-short tau inversion recovery (STIR) compared to the other three contrast images (T1-WI, T2-fluid-attenuated inversion recovery [FLAIR], and T1-FLAIR). In conclusion, PD-WI, PSIR, and PD-STIR should be used instead of T1-WI for skull stripping in the CNN models.

The federated tumor segmentation (FeTS) tool: an open-source solution to further solid tumor research

Objective.De-centralized data analysis becomes an increasingly preferred option in the healthcare domain, as it alleviates the need for sharing primary patient data across collaborating institutions. This highlights the need for consistent harmonized data curation, pre-processing, and identification of regions of interest based on uniform criteria.Approach.Towards this end, this manuscript describes theFederatedTumorSegmentation (FeTS) tool, in terms of software architecture and functionality.Main results.The primary aim of the FeTS tool is to facilitate this harmonized processing and the generation of gold standard reference labels for tumor sub-compartments on brain magnetic resonance imaging, and further enable federated training of a tumor sub-compartment delineation model across numerous sites distributed across the globe, without the need to share patient data.Significance.Building upon existing open-source tools such as the Insight Toolkit and Qt, the FeTS tool is designed to enable training deep learning models targeting tumor delineation in either centralized or federated settings. The target audience of the FeTS tool is primarily the computational researcher interested in developing federated learning models, and interested in joining a global federation towards this effort. The tool is open sourced athttps://github.com/FETS-AI/Front-End.

Expert tumor annotations and radiomics for locally advanced breast cancer in DCE-MRI for ACRIN 6657/I-SPY1

Breast cancer is one of the most pervasive forms of cancer and its inherent intra- and inter-tumor heterogeneity contributes towards its poor prognosis. Multiple studies have reported results from either private institutional data or publicly available datasets. However, current public datasets are limited in terms of having consistency in: a) data quality, b) quality of expert annotation of pathology, and c) availability of baseline results from computational algorithms. To address these limitations, here we propose the enhancement of the I-SPY1 data collection, with uniformly curated data, tumor annotations, and quantitative imaging features. Specifically, the proposed dataset includes a) uniformly processed scans that are harmonized to match intensity and spatial characteristics, facilitating immediate use in computational studies, b) computationally-generated and manually-revised expert annotations of tumor regions, as well as c) a comprehensive set of quantitative imaging (also known as radiomic) features corresponding to the tumor regions. This collection describes our contribution towards repeatable, reproducible, and comparative quantitative studies leading to new predictive, prognostic, and diagnostic assessments.

Enhancing the REMBRANDT MRI collection with expert segmentation labels and quantitative radiomic features

Malignancy of the brain and CNS is unfortunately a common diagnosis. A large subset of these lesions tends to be high grade tumors which portend poor prognoses and low survival rates, and are estimated to be the tenth leading cause of death worldwide. The complex nature of the brain tissue environment in which these lesions arise offers a rich opportunity for translational research. Magnetic Resonance Imaging (MRI) can provide a comprehensive view of the abnormal regions in the brain, therefore, its applications in the translational brain cancer research is considered essential for the diagnosis and monitoring of disease. Recent years has seen rapid growth in the field of radiogenomics, especially in cancer, and scientists have been able to successfully integrate the quantitative data extracted from medical images (also known as radiomics) with genomics to answer new and clinically relevant questions. In this paper, we took raw MRI scans from the REMBRANDT data collection from public domain, and performed volumetric segmentation to identify subregions of the brain. Radiomic features were then extracted to represent the MRIs in a quantitative yet summarized format. This resulting dataset now enables further biomedical and integrative data analysis, and is being made public via the NeuroImaging Tools & Resources Collaboratory (NITRC) repository ( https://www.nitrc.org/projects/rembrandt_brain/ ).

Clinically Deployed Computational Assessment of Multiple Sclerosis Lesions

Multiple Sclerosis (MS) is a demyelinating disease of the central nervous system that affects nearly 1 million adults in the United States. Magnetic Resonance Imaging (MRI) plays a vital role in diagnosis and treatment monitoring in MS patients. In particular, follow-up MRI with T2-FLAIR images of the brain, depicting white matter lesions, is the mainstay for monitoring disease activity and making treatment decisions. In this article, we present a computational approach that has been deployed and integrated into a real-world routine clinical workflow, focusing on two tasks: (a) detecting new disease activity in MS patients, and (b) determining the necessity for injecting Gadolinium Based Contract Agents (GBCAs). This computer-aided detection (CAD) software has been utilized for the former task on more than 19, 000 patients over the course of 10 years, while its added function of identifying patients who need GBCA injection, has been operative for the past 3 years, with > 85% sensitivity. The benefits of this approach are summarized in: (1) offering a reproducible and accurate clinical assessment of MS lesion patients, (2) reducing the adverse effects of GBCAs (and the deposition of GBCAs to the patient's brain) by identifying the patients who may benefit from injection, and (3) reducing healthcare costs, patients' discomfort, and caregivers' workload.

Automated 2D Slice-Based Skull Stripping Multi-View Ensemble Model on NFBS and IBSR Datasets

This study proposed and evaluated a two-dimensional (2D) slice-based multi-view U-Net (MVU-Net) architecture for skull stripping. The proposed model fused all three TI-weighted brain magnetic resonance imaging (MRI) views, i.e., axial, coronal, and sagittal. This 2D method performed equally well as a three-dimensional (3D) model of skull stripping. while using fewer computational resources. The predictions of all three views were fused linearly, producing a final brain mask with better accuracy and efficiency. Meanwhile, two publicly available datasets-the Internet Brain Segmentation Repository (IBSR) and Neurofeedback Skull-stripped (NFBS) repository-were trained and tested. The MVU-Net, U-Net, and skip connection U-Net (SCU-Net) architectures were then compared. For the IBSR dataset, compared to U-Net and SC-UNet, the MVU-Net architecture attained better mean dice score coefficient (DSC), sensitivity, and specificity, at 0.9184, 0.9397, and 0.9908, respectively. Similarly, the MVU-Net architecture achieved better mean DSC, sensitivity, and specificity, at 0.9681, 0.9763, and 0.9954, respectively, than the U-Net and SC-UNet for the NFBS dataset.

Optimization of Deep Learning Based Brain Extraction in MRI for Low Resource Environments

Brain extraction is an indispensable step in neuro-imaging with a direct impact on downstream analyses. Most such methods have been developed for non-pathologically affected brains, and hence tend to suffer in performance when applied on brains with pathologies, e.g., gliomas, multiple sclerosis, traumatic brain injuries. Deep Learning (DL) methodologies for healthcare have shown promising results, but their clinical translation has been limited, primarily due to these methods suffering from i) high computational cost, and ii) specific hardware requirements, e.g., DL acceleration cards. In this study, we explore the potential of mathematical optimizations, towards making DL methods amenable to application in low resource environments. We focus on both the qualitative and quantitative evaluation of such optimizations on an existing DL brain extraction method, designed for pathologically-affected brains and agnostic to the input modality. We conduct direct optimizations and quantization of the trained model (i.e., prior to inference on new data). Our results yield substantial gains, in terms of speedup, latency, through-put, and reduction in memory usage, while the segmentation performance of the initial and the optimized models remains stable, i.e., as quantified by both the Dice Similarity Coefficient and the Hausdorff Distance. These findings support post-training optimizations as a promising approach for enabling the execution of advanced DL methodologies on plain commercial-grade CPUs, and hence contributing to their translation in limited- and low- resource clinical environments.

Reproducibility analysis of multi-institutional paired expert annotations and radiomic features of the Ivy Glioblastoma Atlas Project (Ivy GAP) dataset

PURPOSE

The availability of radiographic magnetic resonance imaging (MRI) scans for the Ivy Glioblastoma Atlas Project (Ivy GAP) has opened up opportunities for development of radiomic markers for prognostic/predictive applications in glioblastoma (GBM). In this work, we address two critical challenges with regard to developing robust radiomic approaches: (a) the lack of availability of reliable segmentation labels for glioblastoma tumor sub-compartments (i.e., enhancing tumor, non-enhancing tumor core, peritumoral edematous/infiltrated tissue) and (b) identifying "reproducible" radiomic features that are robust to segmentation variability across readers/sites.

ACQUISITION AND VALIDATION METHODS

From TCIA's Ivy GAP cohort, we obtained a paired set (n = 31) of expert annotations approved by two board-certified neuroradiologists at the Hospital of the University of Pennsylvania (UPenn) and at Case Western Reserve University (CWRU). For these studies, we performed a reproducibility study that assessed the variability in (a) segmentation labels and (b) radiomic features, between these paired annotations. The radiomic variability was assessed on a comprehensive panel of 11 700 radiomic features including intensity, volumetric, morphologic, histogram-based, and textural parameters, extracted for each of the paired sets of annotations. Our results demonstrated (a) a high level of inter-rater agreement (median value of DICE ≥0.8 for all sub-compartments), and (b) ≈24% of the extracted radiomic features being highly correlated (based on Spearman's rank correlation coefficient) to annotation variations. These robust features largely belonged to morphology (describing shape characteristics), intensity (capturing intensity profile statistics), and COLLAGE (capturing heterogeneity in gradient orientations) feature families.

DATA FORMAT AND USAGE NOTES

We make publicly available on TCIA's Analysis Results Directory (https://doi.org/10.7937/9j41-7d44), the complete set of (a) multi-institutional expert annotations for the tumor sub-compartments, (b) 11 700 radiomic features, and (c) the associated reproducibility meta-analysis.

POTENTIAL APPLICATIONS

The annotations and the associated meta-data for Ivy GAP are released with the purpose of enabling researchers toward developing image-based biomarkers for prognostic/predictive applications in GBM.