Recommendations for quantitative cerebral perfusion MRI using multi-timepoint arterial spin labeling: Acquisition, quantification, and clinical applications

Accurate assessment of cerebral perfusion is vital for understanding the hemodynamic processes involved in various neurological disorders and guiding clinical decision-making. This guidelines article provides a comprehensive overview of quantitative perfusion imaging of the brain using multi-timepoint arterial spin labeling (ASL), along with recommendations for its acquisition and quantification. A major benefit of acquiring ASL data with multiple label durations and/or post-labeling delays (PLDs) is being able to account for the effect of variable arterial transit time (ATT) on quantitative perfusion values and additionally visualize the spatial pattern of ATT itself, providing valuable clinical insights. Although multi-timepoint data can be acquired in the same scan time as single-PLD data with comparable perfusion measurement precision, its acquisition and postprocessing presents challenges beyond single-PLD ASL, impeding widespread adoption. Building upon the 2015 ASL consensus article, this work highlights the protocol distinctions specific to multi-timepoint ASL and provides robust recommendations for acquiring high-quality data. Additionally, we propose an extended quantification model based on the 2015 consensus model and discuss relevant postprocessing options to enhance the analysis of multi-timepoint ASL data. Furthermore, we review the potential clinical applications where multi-timepoint ASL is expected to offer significant benefits. This article is part of a series published by the International Society for Magnetic Resonance in Medicine (ISMRM) Perfusion Study Group, aiming to guide and inspire the advancement and utilization of ASL beyond the scope of the 2015 consensus article.

Synthesizing Contrast-Enhanced MR Images from Noncontrast MR Images Using Deep Learning

BACKGROUND AND PURPOSE

Recent developments in deep learning methods offer a potential solution to the need for alternative imaging methods due to concerns about the toxicity of gadolinium-based contrast agents. The purpose of the study was to synthesize virtual gadolinium contrast-enhanced T1-weighted MR images from noncontrast multiparametric MR images in patients with primary brain tumors by using deep learning.

MATERIALS AND METHODS

We trained and validated a deep learning network by using MR images from 335 subjects in the Brain Tumor Segmentation Challenge 2019 training data set. A held out set of 125 subjects from the Brain Tumor Segmentation Challenge 2019 validation data set was used to test the generalization of the model. A residual inception DenseNet network, called T1c-ET, was developed and trained to simultaneously synthesize virtual contrast-enhanced T1-weighted (vT1c) images and segment the enhancing portions of the tumor. Three expert neuroradiologists independently scored the synthesized vT1c images by using a 3-point Likert scale, evaluating image quality and contrast enhancement against ground truth T1c images (1 = poor, 2 = good, 3 = excellent).

RESULTS

The synthesized vT1c images achieved structural similarity index, peak signal-to-noise ratio, and normalized mean square error scores of 0.91, 64.35, and 0.03, respectively. There was moderate interobserver agreement between the 3 raters, regarding the algorithm's performance in predicting contrast enhancement, with a Fleiss kappa value of 0.61. Our model was able to accurately predict contrast enhancement in 88.8% of the cases (scores of 2 to 3 on the 3-point scale).

CONCLUSIONS

We developed a novel deep learning architecture to synthesize virtual postcontrast enhancement by using only conventional noncontrast brain MR images. Our results demonstrate the potential of deep learning methods to reduce the need for gadolinium contrast in the evaluation of primary brain tumors.

MR Perfusion Imaging of the Lung

Lung perfusion assessment is critical for diagnosing and monitoring a variety of respiratory conditions. MRI perfusion provides a radiation-free technique, making it an ideal choice for longitudinal imaging in younger populations. This review focuses on the techniques and applications of MRI perfusion, including contrast-enhanced (CE) MRI and non-CE methods such as arterial spin labeling (ASL), fourier decomposition (FD), and hyperpolarized 129-Xenon (129-Xe) MRI. ASL leverages endogenous water protons as tracers for a non-invasive measure of lung perfusion, while FD offers simultaneous measurements of lung perfusion and ventilation, enabling the generation of ventilation/perfusion mapsHyperpolarized 129-Xe MRI emerges as a novel tool for assessing regional gas exchange in the lungs. Despite the promise of MRI perfusion techniques, challenges persist, including competition with other imaging techniques and the need for additional validation and standardization. In conditions such as cystic fibrosis and lung cancer, MRI has displayed encouraging results, whereas in diseases like chronic obstructive pulmonary disease, further validation remains necessary. In conclusion, while MRI perfusion techniques hold immense potential for a comprehensive, non-invasive assessment of lung function and perfusion, their broader clinical adoption hinges on technological advancements, collaborative research, and rigorous validation.

Magnetic Resonance Perfusion Imaging for Breast Cancer

Breast cancer is the most frequently diagnosed cancer among women worldwide, carrying a significant socioeconomic burden. Breast cancer is a heterogeneous disease with 4 major subtypes identified. Each subtype has unique prognostic factors, risks, treatment responses, and survival rates. Advances in targeted therapies have considerably improved the 5-year survival rates for primary breast cancer patients largely due to widespread screening programs that enable early detection and timely treatment. Imaging techniques are indispensable in diagnosing and managing breast cancer. While mammography is the primary screening tool, MRI plays a significant role when mammography results are inconclusive or in patients with dense breast tissue. MRI has become standard in breast cancer imaging, providing detailed anatomic and functional data, including tumor perfusion and cellularity. A key characteristic of breast tumors is angiogenesis, a biological process that promotes tumor development and growth. Increased angiogenesis in tumors generally indicates poor prognosis and increased risk of metastasis. Dynamic contrast-enhanced (DCE) MRI measures tumor perfusion and serves as an in vivo metric for angiogenesis. DCE-MRI has become the cornerstone of breast MRI, boasting a high negative-predictive value of 89% to 99%, although its specificity can vary. This review presents a thorough overview of magnetic resonance (MR) perfusion imaging in breast cancer, focusing on the role of DCE-MRI in clinical applications and exploring emerging MR perfusion imaging techniques.

A 3D-printed phantom for quality-controlled reproducibility measurements of arterial spin labeled perfusion

PURPOSE

To develop a portable MR perfusion phantom for quality-controlled assessment and reproducibility of arterial spin labeled (ASL) perfusion measurement.

METHODS

A 3D-printed perfusion phantom was developed that mimics the branching of arterial vessels, capillaries, and a chamber containing cellulose sponge representing tissue characteristics. A peristaltic pump circulated distilled water through the phantom, and was first evaluated at 300, 400, and 500 mL/min. Longitudinal reproducibility of perfusion was performed using 2D pseudo-continuous ASL at 20 post-label delays (PLDs, ranging between 0.2 and 7.8 s at 0.4-s intervals) over a period of 16 weeks, with three repetitions each week. Multi-PLD data were fitted into a general kinetic model for perfusion quantification (f) and arterial transit time (ATT). Intraclass correlation coefficient was used to assess intersession reproducibility.

RESULTS

MR perfusion signals acquired in the 3D-printed perfusion phantom agreed well with the experimental conditions, with progressively increasing signal intensities and decreasing ATT for pump flow rates from 300 to 500 mL/min. The perfusion signal at 400 mL/min and the general kinetic model-derived f and ATT maps were similar across all PLDs for both intrasession and intersession reproducibility. Across all 48 experimental time points, the average f was 75.55 ± 3.83 × 10-3  mL/mL/s, the corresponding ATT was 2.10 ± 0.20 s, and the T1 was 1.84 ± 0.102 s. Intraclass correlation coefficient was 0.92 (95% confidence interval 0.83-0.97) for f, 0.96 (0.91-0.99) for ATT, and 0.94 (0.88-0.98) for T1 , demonstrating excellent reproducibility.

CONCLUSION

A simple, portable 3D-printed perfusion phantom with excellent reproducibility of 2D pseudo-continuous ASL measurements was demonstrated that can serve for quality-controlled and reliable measurements of ASL perfusion.

Synthesizing 3D Multi-Contrast Brain Tumor MRIs Using Tumor Mask Conditioning

Data scarcity and data imbalance are two major challenges in training deep learning models on medical images, such as brain tumor MRI data. The recent advancements in generative artificial intelligence have opened new possibilities for synthetically generating MRI data, including brain tumor MRI scans. This approach can be a potential solution to mitigate the data scarcity problem and enhance training data availability. This work focused on adapting the 2D latent diffusion models to generate 3D multi-contrast brain tumor MRI data with a tumor mask as the condition. The framework comprises two components: a 3D autoencoder model for perceptual compression and a conditional 3D Diffusion Probabilistic Model (DPM) for generating high-quality and diverse multi-contrast brain tumor MRI samples, guided by a conditional tumor mask. Unlike existing works that focused on generating either 2D multi-contrast or 3D single-contrast MRI samples, our models generate multi-contrast 3D MRI samples. We also integrated a conditional module within the UNet backbone of the DPM to capture the semantic class-dependent data distribution driven by the provided tumor mask to generate MRI brain tumor samples based on a specific brain tumor mask. We trained our models using two brain tumor datasets: The Cancer Genome Atlas (TCGA) public dataset and an internal dataset from the University of Texas Southwestern Medical Center (UTSW). The models were able to generate high-quality 3D multi-contrast brain tumor MRI samples with the tumor location aligned by the input condition mask. The quality of the generated images was evaluated using the Fréchet Inception Distance (FID) score. This work has the potential to mitigate the scarcity of brain tumor data and improve the performance of deep learning models involving brain tumor MRI data.

On the application of pseudo-continuous arterial spin labeled MRI for pulmonary perfusion imaging

PURPOSE

To evaluate different approaches for the effective assessment of pulmonary perfusion with a pseudo-continuous arterial spin labeled (pCASL) MRI.

MATERIALS AND METHODS

Four different approaches were evaluated: 1) Cardiac-triggered inferior vena cava (IVC) labeling; 2) IVC labeling with cardiac-triggered acquisition; 3) Right pulmonary artery (RPA) labeling with cardiac-triggered acquisition; and 4) Cardiac-triggered RPA labeling with background suppression (BGS). Each approach was evaluated in 5 healthy volunteers (n = 20) using coefficient of variation (COV) across averages. Approach 4 was also compared against a flow alternating inversion recovery (FAIR).

RESULTS

The IVC labeling (Approach 1) achieved perfusion-weighted images of both lungs, although this approach was more sensitive to variations in heart rate. Cardiac-triggered acquisitions using IVC (Approach 2) and RPA (Approach 3) labeling improved signal consistencies, but were incompatible with BGS. The cardiac-triggered RPA labeling with BGS (Approach 4) achieved a COV of 0.34 ± 0.03 (p < 0.05 compared to IVC labeling approaches) and resulted in perfusion value of 434 ± 64 mL/100 g/min, which was comparable to 451 ± 181 mL/100 g/min measured by FAIR (p = 0.82).

DISCUSSION

Pulmonary perfusion imaging using pCASL-MRI is highly sensitive to cardiac phase, and requires approaches to minimize flow-induced signal variations. Cardiac-triggered RPA labeling with BGS achieves reduced COV and enables robust pulmonary perfusion imaging.

Longitudinal assessment of placental perfusion in normal and hypertensive pregnancies using pseudo-continuous arterial spin-labeled MRI: preliminary experience

OBJECTIVES

To evaluate longitudinal placental perfusion using pseudo-continuous arterial spin-labeled (pCASL) MRI in normal pregnancies and in pregnancies affected by chronic hypertension (cHTN), who are at the greatest risk for placental-mediated disease conditions.

METHODS

Eighteen normal and 23 pregnant subjects with cHTN requiring antihypertensive therapy were scanned at 3 T using free-breathing pCASL-MRI at 16-20 and 24-28 weeks of gestational age.

RESULTS

Mean placental perfusion was 103.1 ± 48.0 and 71.4 ± 18.3 mL/100 g/min at 16-20 and 24-28 weeks respectively in normal pregnancies and 79.4 ± 27.4 and 74.9 ± 26.6 mL/100 g/min in cHTN pregnancies. There was a significant decrease in perfusion between the first and second scans in normal pregnancies (p = 0.004), which was not observed in cHTN pregnancies (p = 0.36). The mean perfusion was not statistically different between normal and cHTN pregnancies at both scans, but the absolute change in perfusion per week was statistically different between these groups (p = 0.044). Furthermore, placental perfusion was significantly lower at both time points (p = 0.027 and 0.044 respectively) in the four pregnant subjects with cHTN who went on to have infants that were small for gestational age (52.7 ± 20.4 and 50.4 ± 20.9 mL/100 g/min) versus those who did not (85 ± 25.6 and 80.0 ± 25.1 mL/100 g/min).

CONCLUSION

pCASL-MRI enables longitudinal assessment of placental perfusion in pregnant subjects. Placental perfusion in the second trimester declined in normal pregnancies whereas it remained unchanged in cHTN pregnancies, consistent with alterations due to vascular disease pathology. Perfusion was significantly lower in those with small for gestational age infants, indicating that pCASL-MRI-measured perfusion may be an effective imaging biomarker for placental insufficiency.

CLINICAL RELEVANCE STATEMENT

pCASL-MRI enables longitudinal assessment of placental perfusion without administering exogenous contrast agent and can identify placental insufficiency in pregnant subjects with chronic hypertension that can lead to earlier interventions.

KEY POINTS

• Arterial spin-labeled (ASL) magnetic resonance imaging (MRI) enables longitudinal assessment of placental perfusion without administering exogenous contrast agent. • ASL-MRI-measured placental perfusion decreased significantly between 16-20 week and 24-28 week gestational age in normal pregnancies, while it remained relatively constant in hypertensive pregnancies, attributed to vascular disease pathology. • ASL-MRI-measured placental perfusion was significantly lower in subjects with hypertension who had a small for gestational age infant at 16-20-week gestation, indicating perfusion as an effective biomarker of placental insufficiency.

MRI-Based Deep Learning Method for Classification of IDH Mutation Status

Isocitrate dehydrogenase (IDH) mutation status has emerged as an important prognostic marker in gliomas. This study sought to develop deep learning networks for non-invasive IDH classification using T2w MR images while comparing their performance to a multi-contrast network. Methods: Multi-contrast brain tumor MRI and genomic data were obtained from The Cancer Imaging Archive (TCIA) and The Erasmus Glioma Database (EGD). Two separate 2D networks were developed using nnU-Net, a T2w-image-only network (T2-net) and a multi-contrast network (MC-net). Each network was separately trained using TCIA (227 subjects) or TCIA + EGD data (683 subjects combined). The networks were trained to classify IDH mutation status and implement single-label tumor segmentation simultaneously. The trained networks were tested on over 1100 held-out datasets including 360 cases from UT Southwestern Medical Center, 136 cases from New York University, 175 cases from the University of Wisconsin-Madison, 456 cases from EGD (for the TCIA-trained network), and 495 cases from the University of California, San Francisco public database. A receiver operating characteristic curve (ROC) was drawn to calculate the AUC value to determine classifier performance. Results: T2-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 85.4% and 87.6% with AUCs of 0.86 and 0.89, respectively. MC-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 91.0% and 92.8% with AUCs of 0.94 and 0.96, respectively. We developed reliable, high-performing deep learning algorithms for IDH classification using both a T2-image-only and a multi-contrast approach. The networks were tested on more than 1100 subjects from diverse databases, making this the largest study on image-based IDH classification to date.

Update on state-of-the-art for arterial spin labeling (ASL) human perfusion imaging outside of the brain

This review article provides an overview of developments for arterial spin labeling (ASL) perfusion imaging in the body (i.e., outside of the brain). It is part of a series of review/recommendation papers from the International Society for Magnetic Resonance in Medicine (ISMRM) Perfusion Study Group. In this review, we focus on specific challenges and developments tailored for ASL in a variety of body locations. After presenting common challenges, organ-specific reviews of challenges and developments are presented, including kidneys, lungs, heart (myocardium), placenta, eye (retina), liver, pancreas, and muscle, which are regions that have seen the most developments outside of the brain. Summaries and recommendations of acquisition parameters (when appropriate) are provided for each organ. We then explore the possibilities for wider adoption of body ASL based on large standardization efforts, as well as the potential opportunities based on recent advances in high/low-field systems and machine-learning. This review seeks to provide an overview of the current state-of-the-art of ASL for applications in the body, highlighting ongoing challenges and solutions that aim to enable more widespread use of the technique in clinical practice.

Topography-based feature extraction of the human placenta from prenatal MR images

Magnetic resonance imaging (MRI) has gained popularity in the field of prenatal imaging due to the ability to provide high quality images of soft tissue. In this paper, we presented a novel method for extracting different textural and morphological features of the placenta from MRI volumes using topographical mapping. We proposed polar and planar topographical mapping methods to produce common placental features from a unique point of observation. The features extracted from the images included the entire placenta surface, as well as the thickness, intensity, and entropy maps displayed in a convenient two-dimensional format. The topography-based images may be useful for clinical placental assessments as well as computer-assisted diagnosis, and prediction of potential pregnancy complications.

Deep learning based automatic segmentation of the placenta and uterine cavity on prenatal MR images

Magnetic resonance imaging (MRI) has potential benefits in understanding fetal and placental complications in pregnancy. An accurate segmentation of the uterine cavity and placenta can help facilitate fast and automated analyses of placenta accreta spectrum and other pregnancy complications. In this study, we trained a deep neural network for fully automatic segmentation of the uterine cavity and placenta from MR images of pregnant women with and without placental abnormalities. The two datasets were axial MRI data of 241 pregnant women, among whom, 101 patients also had sagittal MRI data. Our trained model was able to perform fully automatic 3D segmentation of MR image volumes and achieved an average Dice similarity coefficient (DSC) of 92% for uterine cavity and of 82% for placenta on the sagittal dataset and an average DSC of 87% for uterine cavity and of 82% for placenta on the axial dataset. Use of our automatic segmentation method is the first step in designing an analytics tool for to assess the risk of pregnant women with placenta accreta spectrum.

Corrigendum to: A novel fully automated MRI-based deep-learning method for classification of 1p/19q co-deletion status in brain gliomas

[This corrects the article DOI: 10.1093/noajnl/vdaa066.].

Errata: Brain tumor IDH, 1p/19q, and MGMT molecular classification using MRI-based deep learning: an initial study on the effect of motion and motion correction

[This corrects the article DOI: 10.1117/1.JMI.9.1.016001.].

Recent Technical Developments in ASL: A Review of the State of the Art

This review article provides an overview of a range of recent technical developments in advanced arterial spin labeling (ASL) methods that have been developed or adopted by the community since the publication of a previous ASL consensus paper by Alsop et al. It is part of a series of review/recommendation papers from the International Society for Magnetic Resonance in Medicine Perfusion Study Group. Here, we focus on advancements in readouts and trajectories, image reconstruction, noise reduction, partial volume correction, quantification of nonperfusion parameters, fMRI, fingerprinting, vessel selective ASL, angiography, deep learning, and ultrahigh field ASL. We aim to provide a high level of understanding of these new approaches and some guidance for their implementation, with the goal of facilitating the adoption of such advances by research groups and by MRI vendors. Topics outside the scope of this article that are reviewed at length in separate articles include velocity selective ASL, multiple-timepoint ASL, body ASL, and clinical ASL recommendations.

QU-BraTS: MICCAI BraTS 2020 Challenge on Quantifying Uncertainty in Brain Tumor Segmentation - Analysis of Ranking Scores and Benchmarking Results

Deep learning (DL) models have provided state-of-the-art performance in various medical imaging benchmarking challenges, including the Brain Tumor Segmentation (BraTS) challenges. However, the task of focal pathology multi-compartment segmentation (e.g., tumor and lesion sub-regions) is particularly challenging, and potential errors hinder translating DL models into clinical workflows. Quantifying the reliability of DL model predictions in the form of uncertainties could enable clinical review of the most uncertain regions, thereby building trust and paving the way toward clinical translation. Several uncertainty estimation methods have recently been introduced for DL medical image segmentation tasks. Developing scores to evaluate and compare the performance of uncertainty measures will assist the end-user in making more informed decisions. In this study, we explore and evaluate a score developed during the BraTS 2019 and BraTS 2020 task on uncertainty quantification (QU-BraTS) and designed to assess and rank uncertainty estimates for brain tumor multi-compartment segmentation. This score (1) rewards uncertainty estimates that produce high confidence in correct assertions and those that assign low confidence levels at incorrect assertions, and (2) penalizes uncertainty measures that lead to a higher percentage of under-confident correct assertions. We further benchmark the segmentation uncertainties generated by 14 independent participating teams of QU-BraTS 2020, all of which also participated in the main BraTS segmentation task. Overall, our findings confirm the importance and complementary value that uncertainty estimates provide to segmentation algorithms, highlighting the need for uncertainty quantification in medical image analyses. Finally, in favor of transparency and reproducibility, our evaluation code is made publicly available at https://github.com/RagMeh11/QU-BraTS.

Multi-Site Concordance of Diffusion-Weighted Imaging Quantification for Assessing Prostate Cancer Aggressiveness

BACKGROUND

Diffusion-weighted imaging (DWI) is commonly used to detect prostate cancer, and a major clinical challenge is differentiating aggressive from indolent disease.

PURPOSE

To compare 14 site-specific parametric fitting implementations applied to the same dataset of whole-mount pathologically validated DWI to test the hypothesis that cancer differentiation varies with different fitting algorithms.

STUDY TYPE

Prospective.

POPULATION

Thirty-three patients prospectively imaged prior to prostatectomy.

FIELD STRENGTH/SEQUENCE

3 T, field-of-view optimized and constrained undistorted single-shot DWI sequence.

ASSESSMENT

Datasets, including a noise-free digital reference object (DRO), were distributed to the 14 teams, where locally implemented DWI parameter maps were calculated, including mono-exponential apparent diffusion coefficient (MEADC), kurtosis (K), diffusion kurtosis (DK), bi-exponential diffusion (BID), pseudo-diffusion (BID*), and perfusion fraction (F). The resulting parametric maps were centrally analyzed, where differentiation of benign from cancerous tissue was compared between DWI parameters and the fitting algorithms with a receiver operating characteristic area under the curve (ROC AUC).

STATISTICAL TEST

Levene's test, P < 0.05 corrected for multiple comparisons was considered statistically significant.

RESULTS

The DRO results indicated minimal discordance between sites. Comparison across sites indicated that K, DK, and MEADC had significantly higher prostate cancer detection capability (AUC range = 0.72-0.76, 0.76-0.81, and 0.76-0.80 respectively) as compared to bi-exponential parameters (BID, BID*, F) which had lower AUC and greater between site variation (AUC range = 0.53-0.80, 0.51-0.81, and 0.52-0.80 respectively). Post-processing parameters also affected the resulting AUC, moving from, for example, 0.75 to 0.87 for MEADC varying cluster size.

DATA CONCLUSION

We found that conventional diffusion models had consistent performance at differentiating prostate cancer from benign tissue. Our results also indicated that post-processing decisions on DWI data can affect sensitivity and specificity when applied to radiological-pathological studies in prostate cancer.

LEVEL OF EVIDENCE

1 TECHNICAL EFFICACY: Stage 3.

Automatic Segmentation of Uterine Cavity and Placenta on MR Images Using Deep Learning

Magnetic resonance imaging (MRI) is useful for the detection of abnormalities affecting maternal and fetal health. In this study, we used a fully convolutional neural network for simultaneous segmentation of the uterine cavity and placenta on MR images. We trained the network with MR images of 181 patients, with 157 for training and 24 for validation. The segmentation performance of the algorithm was evaluated using MR images of 60 additional patients that were not involved in training. The average Dice similarity coefficients achieved for the uterine cavity and placenta were 92% and 80%, respectively. The algorithm could estimate the volume of the uterine cavity and placenta with average errors of less than 1.1% compared to manual estimations. Automated segmentation, when incorporated into clinical use, has the potential to quantify, standardize, and improve placental assessment, resulting in improved outcomes for mothers and fetuses.

CascadeNet for hysterectomy prediction in pregnant women due to placenta accreta spectrum

In severe cases, placenta accreta spectrum (PAS) requires emergency hysterectomy, endangering the life of both mother and fetus. Early prediction may reduce complications and aid in management decisions in these high-risk pregnancies. In this work, we developed a novel convolutional network architecture to combine MRI volumes, radiomic features, and custom feature maps to predict PAS severe enough to result in hysterectomy after fetal delivery in pregnant women. We trained, optimized, and evaluated the networks using data from 241 patients, in groups of 157, 24, and 60 for training, validation, and testing, respectively. We found the network using all three paths produced the best performance, with an AUC of 87.8, accuracy 83.3%, sensitivity of 85.0, and specificity of 82.5. This deep learning algorithm, deployed in clinical settings, may identify women at risk before birth, resulting in improved patient outcomes.

Placenta Accreta Spectrum and Hysterectomy Prediction Using MRI Radiomic Features

In women with placenta accreta spectrum (PAS), patient management may involve cesarean hysterectomy at delivery. Magnetic resonance imaging (MRI) has been used for further evaluation of PAS and surgical planning. This work tackles two prediction problems: predicting presence of PAS and predicting hysterectomy using MR images of pregnant patients. First, we extracted approximately 2,500 radiomic features from MR images with two regions of interest: the placenta and the uterus. In addition to analyzing two regions of interest, we dilated the placenta and uterus masks by 5, 10, 15, and 20 mm to gain insights from the myometrium, where the uterus and placenta overlap in the case of PAS. This study cohort includes 241 pregnant women. Of these women, 89 underwent hysterectomy while 152 did not; 141 with suspected PAS, and 100 without suspected PAS. We obtained an accuracy of 0.88 for predicting hysterectomy and an accuracy of 0.92 for classifying suspected PAS. The radiomic analysis tool is further validated, it can be useful for aiding clinicians in decision making on the care of pregnant women.

Brain tumor IDH, 1p/19q, and MGMT molecular classification using MRI-based deep learning: an initial study on the effect of motion and motion correction

Purpose: Deep learning has shown promise for predicting the molecular profiles of gliomas using MR images. Prior to clinical implementation, ensuring robustness to real-world problems, such as patient motion, is crucial. The purpose of this study is to perform a preliminary evaluation on the effects of simulated motion artifact on glioma marker classifier performance and determine if motion correction can restore classification accuracies. Approach: T2w images and molecular information were retrieved from the TCIA and TCGA databases. Simulated motion was added in the k-space domain along the phase encoding direction. Classifier performance for IDH mutation, 1p/19q co-deletion, and MGMT methylation was assessed over the range of 0% to 100% corrupted k-space lines. Rudimentary motion correction networks were trained on the motion-corrupted images. The performance of the three glioma marker classifiers was then evaluated on the motion-corrected images. Results: Glioma marker classifier performance decreased markedly with increasing motion corruption. Applying motion correction effectively restored classification accuracy for even the most motion-corrupted images. For isocitrate dehydrogenase (IDH) classification, 99% accuracy was achieved, exceeding the original performance of the network and representing a new benchmark in non-invasive MRI-based IDH classification. Conclusions: Robust motion correction can facilitate highly accurate deep learning MRI-based molecular marker classification, rivaling invasive tissue-based characterization methods. Motion correction may be able to increase classification accuracy even in the absence of a visible artifact, representing a new strategy for boosting classifier performance.

Deep learning-based segmentation of the placenta and uterus on MR images

Purpose: Magnetic resonance imaging has been recently used to examine the abnormalities of the placenta during pregnancy. Segmentation of the placenta and uterine cavity allows quantitative measures and further analyses of the organs. The objective of this study is to develop a segmentation method with minimal user interaction. Approach: We developed a fully convolutional neural network (CNN) for simultaneous segmentation of the uterine cavity and placenta in three dimensions (3D) while a minimal operator interaction was incorporated for training and testing of the network. The user interaction guided the network to localize the placenta more accurately. In the experiments, we trained two CNNs, one using 70 normal training cases and the other using 129 training cases including normal cases as well as cases with suspected placenta accreta spectrum (PAS). We evaluated the performance of the segmentation algorithms on two test sets: one with 20 normal cases and the other with 50 images from both normal women and women with suspected PAS. Results: For the normal test data, the average Dice similarity coefficient (DSC) was 92% and 82% for the uterine cavity and placenta, respectively. For the combination of normal and abnormal cases, the DSC was 88% and 83% for the uterine cavity and placenta, respectively. The 3D segmentation algorithm estimated the volume of the normal and abnormal uterine cavity and placenta with average volume estimation errors of 4% and 9%, respectively. Conclusions: The deep learning-based segmentation method provides a useful tool for volume estimation and analysis of the placenta and uterus cavity in human placental imaging.

Deciphering Intratumoral Molecular Heterogeneity in Clear Cell Renal Cell Carcinoma with a Radiogenomics Platform

PURPOSE

Intratumoral heterogeneity (ITH) challenges the molecular characterization of clear cell renal cell carcinoma (ccRCC) and is a confounding factor for therapy selection. Most approaches to evaluate ITH are limited by two-dimensional ex vivo tissue analyses. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can noninvasively assess the spatial landscape of entire tumors in their natural milieu. To assess the potential of DCE-MRI, we developed a vertically integrated radiogenomics colocalization approach for multi-region tissue acquisition and analyses. We investigated the potential of spatial imaging features to predict molecular subtypes using histopathologic and transcriptome correlatives.

EXPERIMENTAL DESIGN

We report the results of a prospective study of 49 patients with ccRCC who underwent DCE-MRI prior to nephrectomy. Surgical specimens were sectioned to match the MRI acquisition plane. RNA sequencing data from multi-region tumor sampling (80 samples) were correlated with percent enhancement on DCE-MRI in spatially colocalized regions of the tumor. Independently, we evaluated clinical applicability of our findings in 19 patients with metastatic RCC (39 metastases) treated with first-line antiangiogenic drugs or checkpoint inhibitors.

RESULTS

DCE-MRI identified tumor features associated with angiogenesis and inflammation, which differed within and across tumors, and likely contribute to the efficacy of antiangiogenic drugs and immunotherapies. Our vertically integrated analyses show that angiogenesis and inflammation frequently coexist and spatially anti-correlate in the same tumor. Furthermore, MRI contrast enhancement identifies phenotypes with better response to antiangiogenic therapy among patients with metastatic RCC.

CONCLUSIONS

These findings have important implications for decision models based on biopsy samples and highlight the potential of more comprehensive imaging-based approaches.

Intrasession Reliability of Arterial Spin-Labeled MRI-Measured Noncontrast Perfusion in Glioblastoma at 3 T

Arterial spin-labeled magnetic resonance imaging can provide quantitative perfusion measurements in the brain and can be potentially used to evaluate therapy response assessment in glioblastoma (GBM). The reliability and reproducibility of this method to measure noncontrast perfusion in GBM, however, are lacking. We evaluated the intrasession reliability of brain and tumor perfusion in both healthy volunteers and patients with GBM at 3 T using pseudocontinuous labeling (pCASL) and 3D turbo spin echo (TSE) using Cartesian acquisition with spiral profile reordering (CASPR). Two healthy volunteers at a single time point and 6 newly diagnosed patients with GBM at multiple time points (before, during, and after chemoradiation) underwent scanning (total, 14 sessions). Compared with 3D GraSE, 3D TSE-CASPR generated cerebral blood flow maps with better tumor-to-normal background tissue contrast and reduced image distortions. The intraclass correlation coefficient between the 2 runs of 3D pCASL with TSE-CASPR was consistently high (≥0.90) across all normal-appearing gray matter (NAGM) regions of interest (ROIs), and was particularly high in tumors (0.98 with 95% confidence interval [CI]: 0.97-0.99). The within-subject coefficients of variation were relatively low in all normal-appearing gray matter regions of interest (3.40%-7.12%), and in tumors (4.91%). Noncontrast perfusion measured using 3D pCASL with TSE-CASPR provided robust cerebral blood flow maps in both healthy volunteers and patients with GBM with high intrasession repeatability at 3 T. This approach can be an appropriate noncontrast and noninvasive quantitative perfusion imaging method for longitudinal assessment of therapy response and management of patients with GBM.

A Fully Automated Deep Learning Network for Brain Tumor Segmentation

We developed a fully automated method for brain tumor segmentation using deep learning; 285 brain tumor cases with multiparametric magnetic resonance images from the BraTS2018 data set were used. We designed 3 separate 3D-Dense-UNets to simplify the complex multiclass segmentation problem into individual binary-segmentation problems for each subcomponent. We implemented a 3-fold cross-validation to generalize the network's performance. The mean cross-validation Dice-scores for whole tumor (WT), tumor core (TC), and enhancing tumor (ET) segmentations were 0.92, 0.84, and 0.80, respectively. We then retrained the individual binary-segmentation networks using 265 of the 285 cases, with 20 cases held-out for testing. We also tested the network on 46 cases from the BraTS2017 validation data set, 66 cases from the BraTS2018 validation data set, and 52 cases from an independent clinical data set. The average Dice-scores for WT, TC, and ET were 0.90, 0.84, and 0.80, respectively, on the 20 held-out testing cases. The average Dice-scores for WT, TC, and ET on the BraTS2017 validation data set, the BraTS2018 validation data set, and the clinical data set were as follows: 0.90, 0.80, and 0.78; 0.90, 0.82, and 0.80; and 0.85, 0.80, and 0.77, respectively. A fully automated deep learning method was developed to segment brain tumors into their subcomponents, which achieved high prediction accuracy on the BraTS data set and on the independent clinical data set. This method is promising for implementation into a clinical workflow.

MRI-Based Deep-Learning Method for Determining Glioma MGMT Promoter Methylation Status

BACKGROUND AND PURPOSE

O6-Methylguanine-DNA methyltransferase (MGMT) promoter methylation confers an improved prognosis and treatment response in gliomas. We developed a deep learning network for determining MGMT promoter methylation status using T2 weighted Images (T2WI) only.

MATERIALS AND METHODS

Brain MR imaging and corresponding genomic information were obtained for 247 subjects from The Cancer Imaging Archive and The Cancer Genome Atlas. One hundred sixty-three subjects had a methylated MGMT promoter. A T2WI-only network (MGMT-net) was developed to determine MGMT promoter methylation status and simultaneous single-label tumor segmentation. The network was trained using 3D-dense-UNets. Three-fold cross-validation was performed to generalize the performance of the networks. Dice scores were computed to determine tumor-segmentation accuracy.

RESULTS

The MGMT-net demonstrated a mean cross-validation accuracy of 94.73% across the 3 folds (95.12%, 93.98%, and 95.12%, [SD, 0.66%]) in predicting MGMT methylation status with a sensitivity and specificity of 96.31% [SD, 0.04%] and 91.66% [SD, 2.06%], respectively, and a mean area under the curve of 0.93 [SD, 0.01]. The whole tumor-segmentation mean Dice score was 0.82 [SD, 0.008].

CONCLUSIONS

We demonstrate high classification accuracy in predicting MGMT promoter methylation status using only T2WI. Our network surpasses the sensitivity, specificity, and accuracy of histologic and molecular methods. This result represents an important milestone toward using MR imaging to predict prognosis and treatment response.

Single-shot RARE with Dixon: Application to robust abdominal imaging with uniform fat and water separation at 3T

PURPOSE

To develop a true single shot turbo spin echo (SShTSE) acquisition with Dixon for robust T₂ -weighted abdominal imaging with uniform fat and water separation at 3T.

METHODS

The in-phase (IP) and out-of-phase (OP) echoes for Dixon processing were acquired in the same repetition time of a SShTSE using partial echoes. A phase-preserved bi-directional homodyne reconstruction was developed to compensate the partial echo and the partial phase encoding of SShTSE. With IRB approval, the SShTSE-Dixon was compared against standard SShTSE, without and with fat suppression using spectral adiabatic inversion recovery (SPAIR) in 5 healthy volunteers and 5 patients. The SNR and contrast ratio (CR) of spleen to liver were compared among different acquisitions.

RESULTS

The bi-directional homodyne reconstruction successfully minimized ringing artifacts because of partial acquisitions. SShTSE-Dixon achieved uniform fat suppression compared to SShTSE-SPAIR with fat suppression failures of 1/10 versus 10/10 in the axial plane and 0/5 versus 5/5 in the coronal plane, respectively. The SNRs of the liver (12.2 ± 4.9 vs. 11.7 ± 5.2; P = .76) and spleen (25.9 ± 11.6 vs. 23.7 ± 9.7; P = .14) were equivalent between fat-suppressed images (SShTSE-Dixon water-only and SShTSE-SPAIR). The SNRs of liver (14.4 ± 5.7 vs. 13.4 ± 5.0; P = .60) and spleen (26.5 ± 10.1 vs. 25.7 ± 8.5; P = .56) were equivalent between non-fat-suppressed images (SShTSE-Dixon IP and SShTSE). The CRs of spleen to liver were also similar between fat-suppressed images (2.6 ± 0.4 vs. 2.5 ± 0.5; P =.92) and non-fat-suppressed images (2.3 ± 0.6 vs. 2.2 ± 0.4; P =.84).

CONCLUSION

SShTSE-Dixon generates robust abdominal T₂ -weighted images at 3T with and without uniform fat suppression, along with perfectly co-registered fat-only images in a single acquisition.

Magnetic Resonance Imaging Radiomics Analyses for Prediction of High-Grade Histology and Necrosis in Clear Cell Renal Cell Carcinoma: Preliminary Experience

INTRODUCTION

Percutaneous renal mass biopsy results can accurately diagnose clear cell renal cell carcinoma (ccRCC); however, their reliability to determine nuclear grade in larger, heterogeneous tumors is limited. We assessed the ability of radiomics analyses of magnetic resonance imaging (MRI) to predict high-grade (HG) histology in ccRCC.

PATIENTS AND METHODS

Seventy patients with a renal mass underwent 3 T MRI before surgery between August 2012 and August 2017. Tumor length, first-order statistics, and Haralick texture features were calculated on T2-weighted and dynamic contrast-enhanced (DCE) MRI after manual tumor segmentation. After a variable clustering algorithm was applied, tumor length, washout, and all cluster features were evaluated univariably by receiver operating characteristic curves. Three logistic regression models were constructed to assess the predictability of HG ccRCC and then cross-validated.

RESULTS

At univariate analysis, area under the curve values of length, and DCE texture cluster 1 and cluster 3 for diagnosis of HG ccRCC were 0.7 (95% confidence interval [CI], 0.58-0.82, false discovery rate P = .008), 0.72 (95% CI, 0.59-0.84, false discovery rate P = .004), and 0.75 (95% CI, 0.63-0.87, false discovery rate P = .0009), respectively. At multivariable analysis, area under the curve for model 1 (tumor length only), model 2 (length + DCE clusters 3 and 4), and model 3 (DCE cluster 1 and 3) for diagnosis of HG ccRCC were 0.67 (95% CI, 0.54-0.79), 0.82 (95% CI, 0.71-0.92), and 0.81 (95% CI, 0.70-0.91), respectively.

CONCLUSION

Radiomics analysis of MRI images was superior to tumor size for the prediction of HG histology in ccRCC in our cohort.

Assessing reproducibility in Magnetic Resonance (MR) Radiomics features between Deep-Learning segmented and Expert Manual segmented data and evaluating their diagnostic performance in Pregnant Women with suspected Placenta Accreta Spectrum (PAS)

A Deep-Learning (DL) based segmentation tool was applied to a new magnetic resonance imaging dataset of pregnant women with suspected Placenta Accreta Spectrum (PAS). Radiomic features from DL segmentation were compared to those from expert manual segmentation via intraclass correlation coefficients (ICC) to assess reproducibility. An additional imaging marker quantifying the placental location within the uterus (PLU) was included. Features with an ICC > 0.7 were used to build logistic regression models to predict hysterectomy. Of 2059 features, 781 (37.9%) had ICC >0.7. AUC was 0.69 (95% CI 0.63-0.74) for manually segmented data and 0.78 (95% CI 0.73-0.83) for DL segmented data.

A novel fully automated MRI-based deep-learning method for classification of 1p/19q co-deletion status in brain gliomas

BACKGROUND

One of the most important recent discoveries in brain glioma biology has been the identification of the isocitrate dehydrogenase (IDH) mutation and 1p/19q co-deletion status as markers for therapy and prognosis. 1p/19q co-deletion is the defining genomic marker for oligodendrogliomas and confers a better prognosis and treatment response than gliomas without it. Our group has previously developed a highly accurate deep-learning network for determining IDH mutation status using T2-weighted (T2w) MRI only. The purpose of this study was to develop a similar 1p/19q deep-learning classification network.

METHODS

Multiparametric brain MRI and corresponding genomic information were obtained for 368 subjects from The Cancer Imaging Archive and The Cancer Genome Atlas. 1p/19 co-deletions were present in 130 subjects. Two-hundred and thirty-eight subjects were non-co-deleted. A T2w image-only network (1p/19q-net) was developed to perform 1p/19q co-deletion status classification and simultaneous single-label tumor segmentation using 3D-Dense-UNets. Three-fold cross-validation was performed to generalize the network performance. Receiver operating characteristic analysis was also performed. Dice scores were computed to determine tumor segmentation accuracy.

RESULTS

1p/19q-net demonstrated a mean cross-validation accuracy of 93.46% across the 3 folds (93.4%, 94.35%, and 92.62%, SD = 0.8) in predicting 1p/19q co-deletion status with a sensitivity and specificity of 0.90 ± 0.003 and 0.95 ± 0.01, respectively and a mean area under the curve of 0.95 ± 0.01. The whole tumor segmentation mean Dice score was 0.80 ± 0.007.

CONCLUSION

We demonstrate high 1p/19q co-deletion classification accuracy using only T2w MR images. This represents an important milestone toward using MRI to predict glioma histology, prognosis, and response to treatment.

Combining inhomogeneous magnetization transfer and multipoint Dixon acquisition: Potential utility and evaluation

PURPOSE

The recently introduced inhomogeneous magnetization transfer (ihMT) method has predominantly been applied for imaging the central nervous system. Future applications of ihMT, such as in peripheral nerves and muscles, will involve imaging in the vicinity of adipose tissues. This work aims to systematically investigate the partial volume effect of fat on the ihMT signal and to propose an efficient fat-separation method that does not interfere with ihMT measurements.

METHODS

First, the influence of fat on ihMT signal was studied using simulations. Next, the ihMT sequence was combined with a multi-echo Dixon acquisition for fat separation. The sequence was tested in 9 healthy volunteers using a 3T human scanner. The ihMT ratio (ihMTR) values were calculated in regions of interest in the brain and the spinal cord using standard acquisition (no fat saturation), water-only, in-phase, and out-of-phase reconstructions. The values obtained were compared with a standard fat suppression method, spectral presaturation with inversion recovery.

RESULTS

Simulations showed variations in the ihMTR values in the presence of fat, depending on the TEs used. The IhMTR values in the brain and spinal cord derived from the water-only ihMT multi-echo Dixon images were in good agreement with values from the unsuppressed sequence. The ihMT-spectral presaturation with inversion recovery combination resulted in 24%-35% lower ihMTR values compared with the standard non-fat-suppressed acquisition.

CONCLUSION

The presence of fat within a voxel affects the ihMTR calculations. The IhMT multi-echo Dixon method does not compromise the observable ihMT effect and can potentially be used to remove fat influence in ihMT.

Evaluating the Use of rCBV as a Tumor Grade and Treatment Response Classifier Across NCI Quantitative Imaging Network Sites: Part II of the DSC-MRI Digital Reference Object (DRO) Challenge

We have previously characterized the reproducibility of brain tumor relative cerebral blood volume (rCBV) using a dynamic susceptibility contrast magnetic resonance imaging digital reference object across 12 sites using a range of imaging protocols and software platforms. As expected, reproducibility was highest when imaging protocols and software were consistent, but decreased when they were variable. Our goal in this study was to determine the impact of rCBV reproducibility for tumor grade and treatment response classification. We found that varying imaging protocols and software platforms produced a range of optimal thresholds for both tumor grading and treatment response, but the performance of these thresholds was similar. These findings further underscore the importance of standardizing acquisition and analysis protocols across sites and software benchmarking.

Segmentation of uterus and placenta in MR images using a fully convolutional neural network

Segmentation of the uterine cavity and placenta in fetal magnetic resonance (MR) imaging is useful for the detection of abnormalities that affect maternal and fetal health. In this study, we used a fully convolutional neural network for 3D segmentation of the uterine cavity and placenta while a minimal operator interaction was incorporated for training and testing the network. The user interaction guided the network to localize the placenta more accurately. We trained the network with 70 training and 10 validation MRI cases and evaluated the algorithm segmentation performance using 20 cases. The average Dice similarity coefficient was 92% and 82% for the uterine cavity and placenta, respectively. The algorithm could estimate the volume of the uterine cavity and placenta with average errors of 2% and 9%, respectively. The results demonstrate that the deep learning-based segmentation and volume estimation is possible and can potentially be useful for clinical applications of human placental imaging.

A novel fully automated MRI-based deep-learning method for classification of IDH mutation status in brain gliomas

BACKGROUND

Isocitrate dehydrogenase (IDH) mutation status has emerged as an important prognostic marker in gliomas. Currently, reliable IDH mutation determination requires invasive surgical procedures. The purpose of this study was to develop a highly accurate, MRI-based, voxelwise deep-learning IDH classification network using T2-weighted (T2w) MR images and compare its performance to a multicontrast network.

METHODS

Multiparametric brain MRI data and corresponding genomic information were obtained for 214 subjects (94 IDH-mutated, 120 IDH wild-type) from The Cancer Imaging Archive and The Cancer Genome Atlas. Two separate networks were developed, including a T2w image-only network (T2-net) and a multicontrast (T2w, fluid attenuated inversion recovery, and T1 postcontrast) network (TS-net) to perform IDH classification and simultaneous single label tumor segmentation. The networks were trained using 3D Dense-UNets. Three-fold cross-validation was performed to generalize the networks' performance. Receiver operating characteristic analysis was also performed. Dice scores were computed to determine tumor segmentation accuracy.

RESULTS

T2-net demonstrated a mean cross-validation accuracy of 97.14% ± 0.04 in predicting IDH mutation status, with a sensitivity of 0.97 ± 0.03, specificity of 0.98 ± 0.01, and an area under the curve (AUC) of 0.98 ± 0.01. TS-net achieved a mean cross-validation accuracy of 97.12% ± 0.09, with a sensitivity of 0.98 ± 0.02, specificity of 0.97 ± 0.001, and an AUC of 0.99 ± 0.01. The mean whole tumor segmentation Dice scores were 0.85 ± 0.009 for T2-net and 0.89 ± 0.006 for TS-net.

CONCLUSION

We demonstrate high IDH classification accuracy using only T2-weighted MR images. This represents an important milestone toward clinical translation.

Consensus-based technical recommendations for clinical translation of renal ASL MRI

OBJECTIVES

This study aimed at developing technical recommendations for the acquisition, processing and analysis of renal ASL data in the human kidney at 1.5 T and 3 T field strengths that can promote standardization of renal perfusion measurements and facilitate the comparability of results across scanners and in multi-centre clinical studies.

METHODS

An international panel of 23 renal ASL experts followed a modified Delphi process, including on-line surveys and two in-person meetings, to formulate a series of consensus statements regarding patient preparation, hardware, acquisition protocol, analysis steps and data reporting.

RESULTS

Fifty-nine statements achieved consensus, while agreement could not be reached on two statements related to patient preparation. As a default protocol, the panel recommends pseudo-continuous (PCASL) or flow-sensitive alternating inversion recovery (FAIR) labelling with a single-slice spin-echo EPI readout with background suppression and a simple but robust quantification model.

DISCUSSION

This approach is considered robust and reproducible and can provide renal perfusion images of adequate quality and SNR for most applications. If extended kidney coverage is desirable, a 2D multislice readout is recommended. These recommendations are based on current available evidence and expert opinion. Nonetheless they are expected to be updated as more data become available, since the renal ASL literature is rapidly expanding.

Consensus-based technical recommendations for clinical translation of renal ASL MRI

Objectives

This study aimed at developing technical recommendations for the acquisition, processing and analysis of renal ASL data in the human kidney at 1.5 T and 3 T field strengths that can promote standardization of renal perfusion measurements and facilitate the comparability of results across scanners and in multi-centre clinical studies.

Methods

An international panel of 23 renal ASL experts followed a modified Delphi process, including on-line surveys and two in-person meetings, to formulate a series of consensus statements regarding patient preparation, hardware, acquisition protocol, analysis steps and data reporting.

Results

Fifty-nine statements achieved consensus, while agreement could not be reached on two statements related to patient preparation. As a default protocol, the panel recommends pseudo-continuous (PCASL) or flow-sensitive alternating inversion recovery (FAIR) labelling with a single-slice spin-echo EPI readout with background suppression and a simple but robust quantification model.

Discussion

This approach is considered robust and reproducible and can provide renal perfusion images of adequate quality and SNR for most applications. If extended kidney coverage is desirable, a 2D multislice readout is recommended. These recommendations are based on current available evidence and expert opinion. Nonetheless they are expected to be updated as more data become available, since the renal ASL literature is rapidly expanding.

Electronic supplementary material

The online version of this article (10.1007/s10334-019-00800-z) contains supplementary material, which is available to authorized users.

Classification of brain tumor isocitrate dehydrogenase status using MRI and deep learning

Isocitrate dehydrogenase (IDH) mutation status is an important marker in glioma diagnosis and therapy. We propose an automated pipeline for noninvasively predicting IDH status using deep learning and T2-weighted (T2w) magnetic resonance (MR) images with minimal preprocessing (N4 bias correction and normalization to zero mean and unit variance). T2w MR images and genomic data were obtained from The Cancer Imaging Archive dataset for 260 subjects (120 high-grade and 140 low-grade gliomas). A fully automated two-dimensional densely connected model was trained to classify IDH mutation status on 208 subjects and tested on another held-out set of 52 subjects using fivefold cross validation. Data leakage was avoided by ensuring subject separation during the slice-wise randomization. Mean classification accuracy of 90.5% was achieved for each axial slice in predicting the three classes of no tumor, IDH mutated, and IDH wild type. Test accuracy of 83.8% was achieved in predicting IDH mutation status for individual subjects on the test dataset of 52 subjects. We demonstrate a deep learning method to predict IDH mutation status using T2w MRI alone. Radiologic imaging studies using deep learning methods must address data leakage (subject duplication) in the randomization process to avoid upward bias in the reported classification accuracy.

HIF-2 Complex Dissociation, Target Inhibition, and Acquired Resistance with PT2385, a First-in-Class HIF-2 Inhibitor, in Patients with Clear Cell Renal Cell Carcinoma

PURPOSE

The heterodimeric transcription factor HIF-2 is arguably the most important driver of clear cell renal cell carcinoma (ccRCC). Although considered undruggable, structural analyses at the University of Texas Southwestern Medical Center (UTSW, Dallas, TX) identified a vulnerability in the α subunit, which heterodimerizes with HIF1β, ultimately leading to the development of PT2385, a first-in-class inhibitor. PT2385 was safe and active in a first-in-human phase I clinical trial of patients with extensively pretreated ccRCC at UTSW and elsewhere. There were no dose-limiting toxicities, and disease control ≥4 months was achieved in 42% of patients.

PATIENTS AND METHODS

We conducted a prospective companion substudy involving a subset of patients enrolled in the phase I clinical trial at UTSW (n = 10), who were treated at the phase II dose or above, involving multiparametric MRI, blood draws, and serial biopsies for biochemical, whole exome, and RNA-sequencing studies.

RESULTS

PT2385 inhibited HIF-2 in nontumor tissues, as determined by a reduction in erythropoietin levels (a pharmacodynamic marker), in all but one patient, who had the lowest drug concentrations. PT2385 dissociated HIF-2 complexes in ccRCC metastases, and inhibited HIF-2 target gene expression. In contrast, HIF-1 complexes were unaffected. Prolonged PT2385 treatment resulted in the acquisition of resistance, and we identified a gatekeeper mutation (G323E) in HIF2α, which interferes with drug binding and precluded HIF-2 complex dissociation. In addition, we identified an acquired TP53 mutation elsewhere, suggesting a possible alternate mechanism of resistance.

CONCLUSIONS

These findings demonstrate a core dependency on HIF-2 in metastatic ccRCC and establish PT2385 as a highly specific HIF-2 inhibitor in humans. New approaches will be required to target mutant HIF-2 beyond PT2385 or the closely related PT2977 (MK-6482).

Robust pCASL perfusion imaging using a 3D Cartesian acquisition with spiral profile reordering (CASPR)

PURPOSE

To improve the robustness of arterial spin-labeled measured perfusion using a novel Cartesian acquisition with spiral profile reordering (CASPR) 3D turbo spin echo (TSE) in the brain and kidneys.

METHODS

The CASPR view ordering followed a pseudo-spiral trajectory on a Cartesian grid, by sampling the center of k-space at the beginning of each echo train of a segmented 3D TSE acquisition. With institutional review board approval and written informed consent, 14 normal subjects (9 brain and 5 kidneys) were scanned with pCASL perfusion imaging using 3D CASPR and compared against 3D linear TSE (brain and kidneys), the established 2D EPI and 3D gradient and spin echo perfusion (brain), and 2D single-shot turbo spin-echo perfusion (kidneys). The SNR and the quantitative perfusion values were compared among different acquisitions.

RESULTS

3D CASPR TSE achieved robust perfusion across all slices compared to 3D linear TSE in the brain and kidneys. Compared to 2D EPI, 3D CASPR TSE showed higher SNR across the brain (P < 0.01), and exhibited good agreement (36.4 ± 4.7 and 36.9 ± 5.3 mL/100 g/min with 2D EPI and 3D CASPR, respectively), and with 3D gradient and spin echo (27.9 ± 7.2 mL/100 g/min). Compared to a single slice 2D single-shot turbo spin-echo acquisition, 3D CASPR TSE achieved robust perfusion across the entire kidneys in similar scan time with comparable quantified perfusion values (154.1 ± 74.6 and 151.7 ± 70.6 mL/100 g/min with 2D single-shot turbo spin-echo and 3D CASPR, respectively).

CONCLUSION

The CASPR view ordering with 3D TSE achieves robust arterial spin-labeled perfusion in the brain and kidneys because of the sampling of the center of k-space at the beginning of each echo train.

Dynamic contrast-enhanced MRI to predict intratumoral molecular heterogeneity in clear cell renal cell carcinoma

4580

Background: Mutation/inactivation of VHL in clear cell renal cell carcinoma (ccRCC) leads to upregulation of hypoxia inducible factors ( HIFs) and angiogenesis. However, ccRCC is characterized by high intra-tumor heterogeneity (ITH). Random small samples such as those in percutaneous biopsies are likely limited for characterization of molecular alterations in heterogeneous ccRCCs. We hypothesize that whole-tumor dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) is useful to noninvasively identify ITH in ccRCC. Methods: This IRB-approved, prospective, HIPAA-compliant study, included 62 ccRCCs. 3T DCE MRI was obtained prior to nephrectomy. Surgical specimens were sectioned to match MRI acquisition plane. 182 snap frozen samples (49 tumors) and adjacent uninvolved renal parenchyma (URP) were collected. RNA isolations, cDNA library preparation and mRNA sequencing were performed using standard protocols. RNA expression in 81 tumor samples were correlated (Spearman ranked) with % enhancement in a region of interest (ROI) drawn in the same location of the tumor on pre- and 3 different post-contrast DCE MRI phases. Gene function overrepresentation (OR) analyses were done on top positively and negatively correlated genes. False discovery rate (FDR) < 0.1 was considered statistically significant. Results: Principal component analysis of > 20,000 genes indicated distinct gene expression in tumors from URP. Unsupervised clustering showed enrichment of ccA samples (better prognosis) compared to ccB samples (worse prognosis). Importantly, ccA and ccB samples coexisted in 25% of tumors. DCE-MRI % enhancement correlated with expression of > 300 genes (p < 0.003, FDR < 0.1). OR analyses placed angiogenic pathway gene processes and the immune/inflammatory response processes within the top 5 positively- and negatively-correlated gene functions, respectively. HIF2 target genes correlated positively with % enhancement. Conclusions: DCE MRI detects specific molecular signatures and may help overcome the challenges of ITH in ccRCC. Further research is needed to explore the potential role of DCE MRI to assess response to antiangiogenic and immune-based therapies.

Evaluating Multisite rCBV Consistency from DSC-MRI Imaging Protocols and Postprocessing Software Across the NCI Quantitative Imaging Network Sites Using a Digital Reference Object (DRO)

Relative cerebral blood volume (rCBV) cannot be used as a response metric in clinical trials, in part, because of variations in biomarker consistency and associated interpretation across sites, stemming from differences in image acquisition and postprocessing methods (PMs). This study leveraged a dynamic susceptibility contrast magnetic resonance imaging digital reference object to characterize rCBV consistency across 12 sites participating in the Quantitative Imaging Network (QIN), specifically focusing on differences in site-specific imaging protocols (IPs; n = 17), and PMs (n = 19) and differences due to site-specific IPs and PMs (n = 25). Thus, high agreement across sites occurs when 1 managing center processes rCBV despite slight variations in the IP. This result is most likely supported by current initiatives to standardize IPs. However, marked intersite disagreement was observed when site-specific software was applied for rCBV measurements. This study's results have important implications for comparing rCBV values across sites and trials, where variability in PMs could confound the comparison of therapeutic effectiveness and/or any attempts to establish thresholds for categorical response to therapy. To overcome these challenges and ensure the successful use of rCBV as a clinical trial biomarker, we recommend the establishment of qualifying and validating site- and trial-specific criteria for scanners and acquisition methods (eg, using a validated phantom) and the software tools used for dynamic susceptibility contrast magnetic resonance imaging analysis (eg, using a digital reference object where the ground truth is known).

Assessment of intratumor heterogeneity using imaging texture features in clear cell renal cell carcinoma

663

Background: Intratumoral heterogeneity (ITH) relates to aggressiveness in clear cell renal cell carcinoma (ccRCC), the most common and aggressive subtype of kidney cancer. Percutaneous biopsies have high diagnostic accuracy. However, ITH lowers their reliability in larger, heterogeneous tumors. Haralick texture features extracted from a gray level co-occurrence matrix (GLCM) is a robust method to assess intrinsic tumor imaging characteristics. Some of these features, including entropy as a measure of ITH, have recently been used in differentiating malignant from benign tumors in various organs. We aim to understand how tumor entropy extracted from magnetic resonance (MR) imaging correlate with tumor grade (aggressiveness) and gene expression heterogeneity in ccRCC. Methods: This IRB-approved, prospective study included T2-weighted (T2W) and arterial spin labeled (ASL) MR images of 62 patients with ccRCC. The GLCM was constructed for regions-of interest (ROI) within the tumor and 13 Haralick texture features were estimated. Correlations between texture features and tumor grade were evaluated by logistic regression and quantified by the area under the receiving operating characteristic (ROC) curve (AUC). RNA sequencing of 182 tumor samples in 49 resected tumors was performed. Entropy was correlated with standard deviation (SD) of normalized gene expression levels in multiple samples from the same tumor. Spearman correlation (rho) was computed for each gene. False discovery rate q values < 0.05 were considered statistically significant. Results: Entropy was higher in high-grade than low-grade tumors (11.28 ± 0.52 vs. 10.95 ± 0.65) on T2W (q = 0.028) and ASL (10.45 ± 1.15 vs. 9.65 ± 1.29) (q = 0.013). Entropy had an AUC of 0.70 (T2) for high-grade prediction and was weakly correlated with tumor size (R2 = 0.2). Higher T2 and ASL entropy correlated with higher SD of gene expression. Gene ontology analysis of top correlated genes revealed strong enrichment of genes in metabolic processes. Conclusions: Higher MRI entropy predicts high tumor grade and correlates with increased heterogeneity in gene expression of metabolic processes.

Changing Paradigms in Breast Cancer Screening: Abbreviated Breast MRI

Breast magnetic resonance imaging (MRI) is the most sensitive imaging method for breast cancer detection. In this review we discuss the vastly superior performance of MRI compared to traditional breast cancer screening modalities of mammography, tomosynthesis and ultrasound. We discuss an abbreviated breast MRI (AB-MRI) protocol utilizing Dixon sequences which is compliant with American College of Radiology (ACR) guidelines for accreditation of breast MRI but with significantly reduced scan times. Adaptation of such an AB-MRI protocol significantly increases patient throughput and may allow MRI to serve as a stand- alone breast cancer screening tool.

Frequency Offset Corrected Inversion Pulse for B0 and B1 Insensitive Fat Suppression at 3T: Application to MR Neurography of Brachial Plexus

BACKGROUND

The 3D short tau inversion recovery (STIR) sequence is routinely used in clinical MRI to achieve robust fat suppression. However, the performance of the commonly used adiabatic inversion pulse, hyperbolic secant (HS), is compromised in challenging areas with increased B₀ and B₁ inhomogeneities, such as brachial plexus at 3T.

PURPOSE

To demonstrate the frequency offset corrected inversion (FOCI) pulse as an efficient fat suppression STIR pulse with increased robustness to B₀ and B₁ inhomogeneities at 3T, compared to the HS pulse.

STUDY TYPE

Prospective.

SUBJECTS/PHANTOM

Initial evaluation was performed in phantoms and one healthy volunteer by varying the B₁ field, while subsequent comparison was performed in three healthy volunteers and five patients without varying the B₁ .

FIELD STRENGTH/SEQUENCE

3T; 3D TSE-STIR with HS and FOCI pulses.

ASSESSMENT

Brachial plexus images were qualitatively evaluated by two musculoskeletal radiologists independently using a four-point grading scale for fat suppression, shading artifacts, and nerve visualization.

STATISTICAL TEST

The Wilcoxon signed-rank test with P < 0.05 was considered statistically significant.

RESULTS

Simulations and phantom experiments demonstrated broader bandwidth (2.5 kHz vs. 0.83 kHz, increased B₀ robustness) at the same adiabatic threshold and lower adiabatic threshold (5 μT vs. 7 μT at 3.5 ppm, increased B₁ robustness) at the same bandwidth with the FOCI pulse compared to the HS pulse With increased bandwidth, the FOCI pulse achieved robust fat suppression even at 50% of maximum B₁ strength, while the HS pulse required >75% of maximum B₁ strength. Compared to the standard 3D TSE-STIR with HS pulse, the FOCI pulse achieved uniform fat suppression (P < 0.05), better nerve visualization (P < 0.05), and minimal shading artifacts (P < 0.01) in brachial plexus at 3T.

DATA CONCLUSION

The FOCI pulse has increased robustness to B₀ and B₁ inhomogeneities, compared to the HS pulse, and enables uniform fat suppression in brachial plexus at 3T.

LEVEL OF EVIDENCE

1 Techinical Efficacy: Stage 1 J. Magn. Reson. Imaging 2018;48:1104-1111.

Accelerating chemical exchange saturation transfer MRI with parallel blind compressed sensing

PURPOSE

Chemical exchange saturation transfer is a novel and promising MRI contrast method, but it can be time-consuming. Common parallel imaging methods, like SENSE, can lead to reduced quality of CEST. Here, parallel blind compressed sensing (PBCS), combining blind compressed sensing (BCS) and parallel imaging, is evaluated for the acceleration of CEST in brain and breast.

METHODS

The CEST data were collected in phantoms, brain (N = 3), and breast (N = 2). Retrospective Cartesian undersampling was implemented and the reconstruction results of PBCS-CEST were compared with BCS-CEST and k-t sparse-SENSE CEST. The normalized RMSE and the high-frequency error norm were used for quantitative comparison.

RESULTS

In phantom and in vivo brain experiments, the acceleration factor of R = 10 (24 k-space lines) was achieved and in breast R = 5 (30 k-space lines), without compromising the quality of the PBCS-reconstructed magnetization transfer rate asymmetry maps and Z-spectra. Parallel BCS provides better reconstruction quality when compared with BCS, k-t sparse-SENSE, and SENSE methods using the same number of samples. Parallel BCS overperforms BCS, indicating that the inclusion of coil sensitivity improves the reconstruction of the CEST data.

CONCLUSION

The PBCS method accelerates CEST without compromising its quality. Compressed sensing in combination with parallel imaging can provide a valuable alternative to parallel imaging alone for accelerating CEST experiments.