Multiparametric magnetic resonance imaging to differentiate high-grade gliomas and brain metastases

Differentiation of glioma and solitary brain metastasis: a multi-parameter magnetic resonance imaging study using histogram analysis

BACKGROUND

Differentiation of glioma and solitary brain metastasis (SBM), which requires biopsy or multi-disciplinary diagnosis, remains sophisticated clinically. Histogram analysis of MR diffusion or molecular imaging hasn't been fully investigated for the differentiation and may have the potential to improve it.

METHODS

A total of 65 patients with newly diagnosed glioma or metastases were enrolled. All patients underwent DWI, IVIM, and APTW, as well as the T1W, T2W, T2FLAIR, and contrast-enhanced T1W imaging. The histogram features of apparent diffusion coefficient (ADC) from DWI, slow diffusion coefficient (Dslow), perfusion fraction (frac), fast diffusion coefficient (Dfast) from IVIM, and MTRasym@3.5ppm from APTWI were extracted from the tumor parenchyma and compared between glioma and SBM. Parameters with significant differences were analyzed with the logistics regression and receiver operator curves to explore the optimal model and compare the differentiation performance.

RESULTS

Higher ADCkurtosis (P = 0.022), frackurtosis (P<0.001),and fracskewness (P<0.001) were found for glioma, while higher (MTRasym@3.5ppm)10 (P = 0.045), frac10 (P<0.001),frac90 (P = 0.001), fracmean (P<0.001), and fracentropy (P<0.001) were observed for SBM. frackurtosis (OR = 0.431, 95%CI 0.256-0.723, P = 0.002) was independent factor for SBM differentiation. The model combining (MTRasym@3.5ppm)10, frac10, and frackurtosis showed an AUC of 0.857 (sensitivity: 0.857, specificity: 0.750), while the model combined with frac10 and frackurtosis had an AUC of 0.824 (sensitivity: 0.952, specificity: 0.591). There was no statistically significant difference between AUCs from the two models. (Z = -1.14, P = 0.25).

CONCLUSIONS

The frac10 and frackurtosis in enhanced tumor region could be used to differentiate glioma and SBM and (MTRasym@3.5ppm)10 helps improving the differentiation specificity.

Differentiation Between High-Grade Glioma and Brain Metastasis Using Cerebral Perfusion-Related Parameters (Cerebral Blood Volume and Cerebral Blood Flow): A Systematic Review and Meta-Analysis of Perfusion-weighted MRI Techniques

BACKGROUND

Distinguishing high-grade gliomas (HGGs) from brain metastases (BMs) using perfusion-weighted imaging (PWI) remains challenging. PWI offers quantitative measurements of cerebral blood flow (CBF) and cerebral blood volume (CBV), but optimal PWI parameters for differentiation are unclear.

PURPOSE

To compare CBF and CBV derived from PWIs in HGGs and BMs, and to identify the most effective PWI parameters and techniques for differentiation.

STUDY TYPE

Systematic review and meta-analysis.

POPULATION

Twenty-four studies compared CBF and CBV between HGGs (n = 704) and BMs (n = 488).

FIELD STRENGTH/SEQUENCE

Arterial spin labeling (ASL), dynamic susceptibility contrast (DSC), dynamic contrast-enhanced (DCE), and dynamic susceptibility contrast-enhanced (DSCE) sequences at 1.5 T and 3.0 T.

ASSESSMENT

Following the PRISMA guidelines, four major databases were searched from 2000 to 2024 for studies evaluating CBF or CBV using PWI in HGGs and BMs.

STATISTICAL TESTS

Standardized mean difference (SMD) with 95% CIs was used. Risk of bias (ROB) and publication bias were assessed, and I2 statistic was used to assess statistical heterogeneity. A P-value<0.05 was considered significant.

RESULTS

HGGs showed a significant modest increase in CBF (SMD = 0.37, 95% CI: 0.05-0.69) and CBV (SMD = 0.26, 95% CI: 0.01-0.51) compared with BMs. Subgroup analysis based on region, sequence, ROB, and field strength for CBF (HGGs: 375 and BMs: 222) and CBV (HGGs: 493 and BMs: 378) values were conducted. ASL showed a considerable moderate increase (50% overlapping CI) in CBF for HGGs compared with BMs. However, no significant difference was found between ASL and DSC (P = 0.08).

DATA CONCLUSION

ASL-derived CBF may be more useful than DSC-derived CBF in differentiating HGGs from BMs. This suggests that ASL may be used as an alternative to DSC when contrast medium is contraindicated or when intravenous injection is not feasible.

LEVEL OF EVIDENCE: 1

TECHNICAL EFFICACY

Stage 2.

Differentiation of multiple brain metastases and glioblastoma with multiple foci using MRI criteria

OBJECTIVE

Glioblastoma with multiple foci (mGBM) and multiple brain metastases share several common features on magnetic resonance imaging (MRI). A reliable preoperative diagnosis would be of clinical relevance. The aim of this study was to explore the differences and similarities between mGBM and multiple brain metastases on MRI.

METHODS

We performed a retrospective analysis of 50 patients with mGBM and compared them with a cohort of 50 patients with multiple brain metastases (2-10 lesions) histologically confirmed and treated at our department between 2015 and 2020. The following imaging characteristics were analyzed: lesion location, distribution, morphology, (T2-/FLAIR-weighted) connections between the lesions, patterns of contrast agent uptake, apparent diffusion coefficient (ADC)-values within the lesion, the surrounding T2-hyperintensity, and edema distribution.

RESULTS

A total of 210 brain metastases and 181 mGBM lesions were analyzed. An infratentorial localization was found significantly more often in patients with multiple brain metastases compared to mGBM patients (28 vs. 1.5%, p < 0.001). A T2-connection between the lesions was detected in 63% of mGBM lesions compared to 1% of brain metastases. Cortical edema was only present in mGBM. Perifocal edema with larger areas of diffusion restriction was detected in 31% of mGBM patients, but not in patients with metastases.

CONCLUSION

We identified a set of imaging features which improve preoperative diagnosis. The presence of T2-weighted imaging hyperintensity connection between the lesions and cortical edema with varying ADC-values was typical for mGBM.

Machine learning decision tree models for multiclass classification of common malignant brain tumors using perfusion and spectroscopy MRI data

Background

To investigate the contribution of machine learning decision tree models applied to perfusion and spectroscopy MRI for multiclass classification of lymphomas, glioblastomas, and metastases, and then to bring out the underlying key pathophysiological processes involved in the hierarchization of the decision-making algorithms of the models.

Methods

From 2013 to 2020, 180 consecutive patients with histopathologically proved lymphomas (n = 77), glioblastomas (n = 45), and metastases (n = 58) were included in machine learning analysis after undergoing MRI. The perfusion parameters (rCBVmax, PSRmax) and spectroscopic concentration ratios (lac/Cr, Cho/NAA, Cho/Cr, and lip/Cr) were applied to construct Classification and Regression Tree (CART) models for multiclass classification of these brain tumors. A 5-fold random cross validation was performed on the dataset.

Results

The decision tree model thus constructed successfully classified all 3 tumor types with a performance (AUC) of 0.98 for PCNSLs, 0.98 for GBM and 1.00 for METs. The model accuracy was 0.96 with a RSquare of 0.887. Five rules of classifier combinations were extracted with a predicted probability from 0.907 to 0.989 for that end nodes of the decision tree for tumor multiclass classification. In hierarchical order of importance, the root node (Cho/NAA) in the decision tree algorithm was primarily based on the proliferative, infiltrative, and neuronal destructive characteristics of the tumor, the internal node (PSRmax), on tumor tissue capillary permeability characteristics, and the end node (Lac/Cr or Cho/Cr), on tumor energy glycolytic (Warburg effect), or on membrane lipid tumor metabolism.

Conclusion

Our study shows potential implementation of machine learning decision tree model algorithms based on a hierarchical, convenient, and personalized use of perfusion and spectroscopy MRI data for multiclass classification of these brain tumors.

Conventional and Advanced Magnetic Resonance Imaging Assessment of Non-Enhancing Peritumoral Area in Brain Tumor

The non-enhancing peritumoral area (NEPA) is defined as the hyperintense region in T2-weighted and fluid-attenuated inversion recovery (FLAIR) images surrounding a brain tumor. The NEPA corresponds to different pathological processes, including vasogenic edema and infiltrative edema. The analysis of the NEPA with conventional and advanced magnetic resonance imaging (MRI) was proposed in the differential diagnosis of solid brain tumors, showing higher accuracy than MRI evaluation of the enhancing part of the tumor. In particular, MRI assessment of the NEPA was demonstrated to be a promising tool for distinguishing high-grade gliomas from primary lymphoma and brain metastases. Additionally, the MRI characteristics of the NEPA were found to correlate with prognosis and treatment response. The purpose of this narrative review was to describe MRI features of the NEPA obtained with conventional and advanced MRI techniques to better understand their potential in identifying the different characteristics of high-grade gliomas, primary lymphoma and brain metastases and in predicting clinical outcome and response to surgery and chemo-irradiation. Diffusion and perfusion techniques, such as diffusion tensor imaging (DTI), diffusional kurtosis imaging (DKI), dynamic susceptibility contrast-enhanced (DSC) perfusion imaging, dynamic contrast-enhanced (DCE) perfusion imaging, arterial spin labeling (ASL), spectroscopy and amide proton transfer (APT), were the advanced MRI procedures we reviewed.

Differentiation of high grade glioma and solitary brain metastases by measuring relative cerebral blood volume and fractional anisotropy: a systematic review and meta-analysis of MRI diagnostic test accuracy studies

OBJECTIVE

This study aims to research the efficacy of MRI (I) for differentiating high-grade glioma (HGG) (P) with solitary brain metastasis (SBM) (C) by creating a combination of relative cerebral blood volume (rCBV) (O) and fractional anisotropy (FA) (O) in patients with intracerebral tumors.

METHODS

Searches were conducted on September 2021 with no publication date restriction, using an electronic search for related articles published in English, from PubMed (1994 to September 2021), Scopus (1977 to September 2021), Web of Science (1985 to September 2021), and Cochrane (1997 to September 2021). A total of 1056 studies were found, with 23 used for qualitative and quantitative data synthesis. Inclusion criteria were: patients diagnosed with HGG and SBM without age, sex, or race restriction; MRI examination of rCBV and FA; reliable histopathological diagnostic method as the gold-standard for all conditions of interest; observational and clinical studies. Newcastle-Ottawa quality assessment Scale (NOS) and Cochrane risk of bias tool (ROB) for observational and clinical trial studies were managed to appraise the quality of individual studies included. Data extraction results were managed using Mendeley and Excel, pooling data synthesis was completed using the Review Manager 5.4 software with random effect model to discriminate HGG and SBM, and divided into four subgroups.

RESULTS

There were 23 studies included with a total sample size of 597 HGG patients and 373 control groups/SBM. The analysis was categorized into four subgroups: (1) the subgroup with rCBV values in the central area of the tumor/intratumoral (399 HGG and 232 SBM) shows that HGG patients are not significantly different from SBM/controls group (SMD [95% CI] = -0.27 [-0.66, 0.13]), 2) the subgroup with rCBV values in the peritumoral area (452 HGG and 274 SBM) shows that HGG patients are significantly higher than SBM (SMD [95% CI] =  -1.23 [-1.45 to -1.01]), (3) the subgroup with FA values in the central area of the tumor (249 HGG and 156 SBM) shows that HGG patients are significantly higher than SBM (SMD [95% CI] = - 0.44 [-0.84,-0.04]), furthermore (4) the subgroup with FA values in the peritumoral area (261 HGG and 168 SBM) shows that the HGG patients are significantly higher than the SBM (SMD [95% CI] = -0.59 [-1.02,-0.16]).

CONCLUSION

Combining rCBV and FA measurements in the peritumoral region and FA in the intratumoral region increase the accuracy of MRI examination to differentiate between HGG and SBM patients effectively. Confidence in the accuracy of our results may be influenced by major interstudy heterogeneity. Whereas the I2 for the rCBV in the intratumoral subgroup was 80%, I2 for the rCBV in the peritumoral subgroup was 39%, and I2 for the FA in the intratumoral subgroup was 69%, and I2 for the FA in the peritumoral subgroup was 74%. The predefined accurate search criteria, and precise selection and evaluation of methodological quality for included studies, strengthen this studyOur study has no funder, no conflict of interest, and followed an established PROSPERO protocol (ID: CRD42021279106).

ADVANCES IN KNOWLEDGE

The combination of rCBV and FA measurements' results is promising in differentiating HGG and SBM.

Artificial intelligence in multi-parametric magnetic resonance imaging: A review

Multi-parametric magnetic resonance imaging (mpMRI) is an indispensable tool in the clinical workflow for the diagnosis and treatment planning of various diseases. Machine learning-based artificial intelligence (AI) methods, especially those adopting the deep learning technique, have been extensively employed to perform mpMRI image classification, segmentation, registration, detection, reconstruction, and super-resolution. The current availability of increasing computational power and fast-improving AI algorithms have empowered numerous computer-based systems for applying mpMRI to disease diagnosis, imaging-guided radiotherapy, patient risk and overall survival time prediction, and the development of advanced quantitative imaging technology for magnetic resonance fingerprinting. However, the wide application of these developed systems in the clinic is still limited by a number of factors, including robustness, reliability, and interpretability. This survey aims to provide an overview for new researchers in the field as well as radiologists with the hope that they can understand the general concepts, main application scenarios, and remaining challenges of AI in mpMRI. This article is protected by copyright. All rights reserved.

Differentiating solitary brain metastases from high-grade gliomas with MR: comparing qualitative versus quantitative diagnostic strategies

PURPOSE

To investigate the diagnostic efficacy of MRI diagnostic algorithms with an ascending automatization, in distinguishing between high-grade glioma (HGG) and solitary brain metastases (SBM).

METHODS

36 patients with histologically proven HGG (n = 18) or SBM (n = 18), matched by size and location were enrolled from a database containing 655 patients. Four different diagnostic algorithms were performed serially to mimic the clinical setting where a radiologist would typically seek out further findings to reach a decision: pure qualitative, analytic qualitative (based on standardized evaluation of tumor features), semi-quantitative (based on perfusion and diffusion cutoffs included in the literature) and a quantitative data-driven algorithm of the perfusion and diffusion parameters. The diagnostic yields of the four algorithms were tested with ROC analysis and Kendall coefficient of concordance.

RESULTS

Qualitative algorithm yielded sensitivity of 72.2%, specificity of 78.8%, and AUC of 0.75. Analytic qualitative algorithm distinguished HGG from SBM with a sensitivity of 100%, specificity of 77.7%, and an AUC of 0.889. The semi-quantitative algorithm yielded sensitivity of 94.4%, specificity of 83.3%, and AUC = 0.889. The data-driven algorithm yielded sensitivity = 94.4%, specificity = 100%, and AUC = 0.948. The concordance analysis between the four algorithms and the histologic findings showed moderate concordance for the first algorithm, (k = 0.501, P < 0.01), good concordance for the second (k = 0.798, P < 0.01), and third (k = 0.783, P < 0.01), and excellent concordance for fourth (k = 0.901, p < 0.0001).

CONCLUSION

When differentiating HGG from SBM, an analytical qualitative algorithm outperformed qualitative algorithm, and obtained similar results compared to the semi-quantitative approach. However, the use of data-driven quantitative algorithm yielded an excellent differentiation.

Image-Based Differentiation of Intracranial Metastasis From Glioblastoma Using Automated Machine Learning

Purpose

The majority of solitary brain metastases appear similar to glioblastomas (GBMs) on magnetic resonance imaging (MRI). This study aimed to develop and validate an MRI-based model to differentiate intracranial metastases from GBMs using automated machine learning.

Materials and Methods

Radiomics features from 354 patients with brain metastases and 354 with GBMs were used to build prediction algorithms based on T2-weighted images, contrast-enhanced (CE) T1-weighted images, or both. The data of these subjects were subjected to a nested 10-fold split in the training and testing groups to build the best algorithms using the tree-based pipeline optimization tool (TPOT). The algorithms were independently validated using data from 124 institutional patients with solitary brain metastases and 103 patients with GBMs from the cancer genome atlas.

Results

Three groups of models were developed. The average areas under the receiver operating characteristic curve (AUCs) were 0.856 for CE T1-weighted images, 0.976 for T2-weighted images, and 0.988 for a combination in the testing groups, and the AUCs of the groups of models in the independent validation were 0.687, 0.831, and 0.867, respectively. A total of 149 radiomics features were considered as the most valuable features for the differential diagnosis of GBMs and metastases.

Conclusion

The models established by TPOT can distinguish glioblastoma from solitary brain metastases well, and its non-invasiveness, convenience, and robustness make it potentially useful for clinical applications.

Single brain metastasis versus glioblastoma multiforme: a VOI-based multiparametric analysis for differential diagnosis

Purpose

The authors’ purpose was to create a valid multiparametric MRI model for the differential diagnosis between glioblastoma and solitary brain metastasis.

Materials and methods

Forty-one patients (twenty glioblastomas and twenty-one brain metastases) were retrospectively evaluated. MRIs were analyzed with Olea Sphere^® 3.0. Lesions’ volumes of interest (VOIs) were drawn on enhanced 3D T1 MP-RAGE and projected on ADC and rCBV co-registered maps. Another two VOIs were drawn in the region of hyperintense cerebral edema, surrounding the lesion, respectively, within 5 mm around the enhancing tumor and into residual edema. Perfusion curves were obtained, and the value of signal recovery (SR) was reported. A two-sample T test was obtained to compare all parameters of GB and BM groups. Receiver operating characteristics (ROC) analysis was performed.

Results

According to ROC analysis, the area under the curve was 88%, 78% and 74%, respectively, for mean ADC VOI values of the solid component, the mean and max rCBV values in the perilesional edema and the PSR. The cumulative ROC curve of these parameters reached an area under the curve of 95%. Using perilesional max rCBV > 1.37, PSR > 75% and mean lesional ADC < 1 × 10⁻³ mm² s⁻¹ GB could be differentiated from solitary BM (sensitivity and specificity of 95% and 86%).

Conclusion

Lower values of ADC in the enhancing tumor, a higher percentage of SR in perfusion curves and higher values of rCBV in the peritumoral edema closed to the lesion are strongly indicative of GB than solitary BM.

Development of a Machine Learning Classifier Based on Radiomic Features Extracted From Post-Contrast 3D T1-Weighted MR Images to Distinguish Glioblastoma From Solitary Brain Metastasis

Objectives

To differentiate Glioblastomas (GBM) and Brain Metastases (BM) using a radiomic features-based Machine Learning (ML) classifier trained from post-contrast three-dimensional T1-weighted (post-contrast 3DT1) MR imaging, and compare its performance in medical diagnosis versus human experts, on a testing cohort.

Methods

We enrolled 143 patients (71 GBM and 72 BM) in a retrospective bicentric study from January 2010 to May 2019 to train the classifier. Post-contrast 3DT1 MR images were performed on a 3-Tesla MR unit and 100 radiomic features were extracted. Selection and optimization of the Machine Learning (ML) classifier was performed using a nested cross-validation. Sensitivity, specificity, balanced accuracy, and area under the receiver operating characteristic curve (AUC) were calculated as performance metrics. The model final performance was cross-validated, then evaluated on a test set of 37 patients, and compared to human blind reading using a McNemar’s test.

Results

The ML classifier had a mean [95% confidence interval] sensitivity of 85% [77; 94], a specificity of 87% [78; 97], a balanced accuracy of 86% [80; 92], and an AUC of 92% [87; 97] with cross-validation. Sensitivity, specificity, balanced accuracy and AUC were equal to 75, 86, 80 and 85% on the test set. Sphericity 3D radiomic index highlighted the highest coefficient in the logistic regression model. There were no statistical significant differences observed between the performance of the classifier and the experts’ blinded examination.

Conclusions

The proposed diagnostic support system based on radiomic features extracted from post-contrast 3DT1 MR images helps in differentiating solitary BM from GBM with high diagnosis performance and generalizability.

Machine learning based differentiation of glioblastoma from brain metastasis using MRI derived radiomics

Few studies have addressed radiomics based differentiation of Glioblastoma (GBM) and intracranial metastatic disease (IMD). However, the effect of different tumor masks, comparison of single versus multiparametric MRI (mp-MRI) or select combination of sequences remains undefined. We cross-compared multiple radiomics based machine learning (ML) models using mp-MRI to determine optimized configurations. Our retrospective study included 60 GBM and 60 IMD patients. Forty-five combinations of ML models and feature reduction strategies were assessed for features extracted from whole tumor and edema masks using mp-MRI [T1W, T2W, T1-contrast enhanced (T1-CE), ADC, FLAIR], individual MRI sequences and combined T1-CE and FLAIR sequences. Model performance was assessed using receiver operating characteristic curve. For mp-MRI, the best model was LASSO model fit using full feature set (AUC 0.953). FLAIR was the best individual sequence (LASSO-full feature set, AUC 0.951). For combined T1-CE/FLAIR sequence, adaBoost-full feature set was the best performer (AUC 0.951). No significant difference was seen between top models across all scenarios, including models using FLAIR only, mp-MRI and combined T1-CE/FLAIR sequence. Top features were extracted from both the whole tumor and edema masks. Shape sphericity is an important discriminating feature.

Texture Analysis in Brain Tumor MR Imaging

Texture analysis, as well as its broader category radiomics, describes a variety of techniques for image analysis that quantify the variation in surface intensity or patterns, including some that are imperceptible to the human visual system. Cerebral gliomas have been most rigorously studied in brain tumors using MR-based texture analysis (MRTA) to determine the correlation of various clinical measures with MRTA features. Promising results in cerebral gliomas have been shown in the previous MRTA studies in terms of the correlation with the World Health Organization grades, risk stratification in gliomas, and the differentiation of gliomas from other brain tumors. Multiple MRTA studies in gliomas have repeatedly shown high performance of entropy, a measure of the randomness in image intensity values, of either histogram- or gray-level co-occurrence matrix parameters. Similarly, researchers have applied MRTA to other brain tumors, including meningiomas and pediatric posterior fossa tumors.

However, the value of MRTA in the clinical use remains undetermined, probably because previous studies have shown only limited reproducibility of the result in the real world. The low-to-modest generalizability may be attributed to variations in MRTA methods, sampling bias that originates from single-institution studies, and overfitting problems to a limited number of samples.

To enhance the reliability and reproducibility of MRTA studies, researchers have realized the importance of standardizing methods in the field of radiomics. Another advancement is the recent development of a comprehensive assessment system to ensure the quality of a radiomics study. These two-way approaches will secure the validity of upcoming MRTA studies. The clinical use of texture analysis in brain MRI will be accelerated by these continuous efforts.

Radiomic analysis of magnetic resonance fingerprinting in adult brain tumors

PURPOSE

This is a radiomics study investigating the ability of texture analysis of MRF maps to improve differentiation between intra-axial adult brain tumors and to predict survival in the glioblastoma cohort.

METHODS

Magnetic resonance fingerprinting (MRF) acquisition was performed on 31 patients across 3 groups: 17 glioblastomas, 6 low-grade gliomas, and 8 metastases. Using regions of interest for the solid tumor and peritumoral white matter on T1 and T2 maps, second-order texture features were calculated from gray-level co-occurrence matrices and gray-level run length matrices. Selected features were compared across the three tumor groups using Wilcoxon rank-sum test. Receiver operating characteristic curve analysis was performed for each feature. Kaplan-Meier method was used for survival analysis with log rank tests.

RESULTS

Low-grade gliomas and glioblastomas had significantly higher run percentage, run entropy, and information measure of correlation 1 on T1 than metastases (p < 0.017). The best separation of all three tumor types was seen utilizing inverse difference normalized and homogeneity values for peritumoral white matter in both T1 and T2 maps (p < 0.017). In solid tumor T2 maps, lower values in entropy and higher values of maximum probability and high-gray run emphasis were associated with longer survival in glioblastoma patients (p < 0.05). Several texture features were associated with longer survival in glioblastoma patients on peritumoral white matter T1 maps (p < 0.05).

CONCLUSION

Texture analysis of MRF-derived maps can improve our ability to differentiate common adult brain tumors by characterizing tumor heterogeneity, and may have a role in predicting outcomes in patients with glioblastoma.

Differentiating High-Grade Gliomas from Brain Metastases at Magnetic Resonance: The Role of Texture Analysis of the Peritumoral Zone

High-grade gliomas (HGGs) and solitary brain metastases (BMs) have similar imaging appearances, which often leads to misclassification. In HGGs, the surrounding tissues show malignant invasion, while BMs tend to displace the adjacent area. The surrounding edema produced by the two cannot be differentiated by conventional magnetic resonance (MRI) examinations. Forty-two patients with pathology-proven brain tumors who underwent conventional pretreatment MRIs were retrospectively included (HGGs, n = 16; BMs, n = 26). Texture analysis of the peritumoral zone was performed on the T2-weighted sequence using dedicated software. The most discriminative texture features were selected using the Fisher and the probability of classification error and average correlation coefficients. The ability of texture parameters to distinguish between HGGs and BMs was evaluated through univariate, receiver operating, and multivariate analyses. The first percentile and wavelet energy texture parameters were independent predictors of HGGs (75–87.5% sensitivity, 53.85–88.46% specificity). The prediction model consisting of all parameters that showed statistically significant results at the univariate analysis was able to identify HGGs with 100% sensitivity and 66.7% specificity. Texture analysis can provide a quantitative description of the peritumoral zone encountered in solitary brain tumors, that can provide adequate differentiation between HGGs and BMs.

Next-Generation Sequencing Analysis of ctDNA for the Detection of Glioma and Metastatic Brain Tumors in Adults

Background and aims: The next-generation sequencing technologies and their related assessments of circulating tumor DNA in both glioma and metastatic brain tumors remain largely limited.

Methods: Based largely on a protocol approved by the institutional review board at Peking University International Hospital, the current retrospective, single-center study was conducted. Genomic DNA was extracted from blood samples or tumor tissues. With the application of NextSeq 500 instrument (Illumina), Sequencing was performed with an average coverage of 550-fold. Paired-end sequencing was employed utilized with an attempt to achieve improved sensitivity of duplicate detection and therefore to increase the detection reliability of possible fusions.

Results: A total of 28 patients (21 men and 7 women) with brain tumors in the present study were involved in the study. The patients enrolled were assigned into two groups, including glioma group (n = 21) and metastatic brain tumor group (n = 7). The mean age of metastatic brain tumor group (59.86 ± 8.85 y), (43.65 ± 13.05 y) reported significantly higher results in comparison to that of glioma group (45.3 ± 12.3 years) (P < 0.05). The mutant genes in metastatic brain tumor group included ALK, MDM2, ATM, BRCA1, FGFR1, MDM4 and KRAS; however, there were no glioma-related mutant genes including MGMT, IDH1, IDH2, 1p/19q, and BRAF et al. Interesteringly, only two patient (28.3%) was detected blood ctDNA in metastatic brain tumor group; In contrast, blood ctDNA was found in ten glioma patients (47.6%) including 1p/19q, MDM2, ERBB2, IDH1, CDKN2A, CDK4, PDGFRA, CCNE1, MET. The characterizations of IDH mutations in the glioma included IDH1 mutation (p.R132H) and IDH2 mutation (p.R172K). The mutation rate of IDH in tumor tissues was 37.06 ± 8.32%, which was significantly higher than blood samples (P < 0.05).

Conclusion: The present study demonstrated that the mutant genes among glioma and metastatic brain tumors are shown to be different. Moreover, the ctDNAs in the metastatic brain tumors included ALK and MDM2, and glioma-related ctDNAs included 1p/19q and MDM2 followed by frequencies of ERBB2, IDH1, CDKN2A, CDK4, PDGFRA, CCNE1, MET. These ctDNAs might be biomarkers and therapeutic responders in brain tumor.

Automated glioma grading on conventional MRI images using deep convolutional neural networks

PURPOSE

Gliomas are the most common primary tumor of the brain and are classified into grades I-IV of the World Health Organization (WHO), based on their invasively histological appearance. Gliomas grading plays an important role to determine the treatment plan and prognosis prediction. In this study we propose two novel methods for automatic, non-invasively distinguishing low-grade (Grades II and III) glioma (LGG) and high-grade (grade IV) glioma (HGG) on conventional MRI images by using deep convolutional neural networks (CNNs).

METHODS

All MRI images have been preprocessed first by rigid image registration and intensity inhomogeneity correction. Both proposed methods consist of two steps: (a) three-dimensional (3D) brain tumor segmentation based on a modification of the popular U-Net model; (b) tumor classification on segmented brain tumor. In the first method, the slice with largest area of tumor is determined and the state-of-the-art mask R-CNN model is employed for tumor grading. To improve the performance of the grading model, a two-dimensional (2D) data augmentation has been implemented to increase both the amount and the diversity of the training images. In the second method, denoted as 3DConvNet, a 3D volumetric CNNs is applied directly on bounding image regions of segmented tumor for classification, which can fully leverage the 3D spatial contextual information of volumetric image data.

RESULTS

The proposed schemes were evaluated on The Cancer Imaging Archive (TCIA) low grade glioma (LGG) data, and the Multimodal Brain Tumor Image Segmentation (BraTS) Benchmark 2018 training datasets with fivefold cross validation. All data are divided into training, validation, and test sets. Based on biopsy-proven ground truth, the performance metrics of sensitivity, specificity, and accuracy are measured on the test sets. The results are 0.935 (sensitivity), 0.972 (specificity), and 0.963 (accuracy) for the 2D Mask R-CNN based method, and 0.947 (sensitivity), 0.968 (specificity), and 0.971 (accuracy) for the 3DConvNet method, respectively. In regard to efficiency, for 3D brain tumor segmentation, the program takes around ten and a half hours for training with 300 epochs on BraTS 2018 dataset and takes only around 50 s for testing of a typical image with a size of 160 × 216 × 176. For 2D Mask R-CNN based tumor grading, the program takes around 4 h for training with around 60 000 iterations, and around 1 s for testing of a 2D slice image with size of 128 × 128. For 3DConvNet based tumor grading, the program takes around 2 h for training with 10 000 iterations, and 0.25 s for testing of a 3D cropped image with size of 64 × 64 × 64, using a DELL PRECISION Tower T7910, with two NVIDIA Titan Xp GPUs.

CONCLUSIONS

Two effective glioma grading methods on conventional MRI images using deep convolutional neural networks have been developed. Our methods are fully automated without manual specification of region-of-interests and selection of slices for model training, which are common in traditional machine learning based brain tumor grading methods. This methodology may play a crucial role in selecting effective treatment options and survival predictions without the need for surgical biopsy.

Glioblastomas and brain metastases differentiation following an MRI texture analysis-based radiomics approach

PURPOSE

To evaluate the potential of 2D texture features extracted from magnetic resonance (MR) images for differentiating brain metastasis (BM) and glioblastomas (GBM) following a radiomics approach.

METHODS

This retrospective study included 50 patients with BM and 50 with GBM who underwent T1-weighted MRI between December 2010 and January 2017. Eighty-eight rotation-invariant texture features were computed for each segmented lesion using six texture analysis methods. These features were also extracted from the four images obtained after applying the discrete wavelet transform (88 features × 4 images). Three feature selection methods and five predictive models were evaluated. A 5-fold cross-validation scheme was used to randomly split the study group into training (80 patients) and testing (20 patients), repeating the process ten times. Classification was evaluated computing the average area under the receiver operating characteristic curve. Sensibility, specificity and accuracy were also computed. The whole process was tested quantizing the images with different gray-level values to evaluate their influence in the final results.

RESULTS

Highest classification accuracy was obtained using the original images quantized with 128 gray-levels and a feature selection method based on the p-value. The best overall performance was achieved using a support vector machine model with a subset of 32 features (AUC = 0.896 ± 0.067, sensitivity of 82% and specificity of 80%). Naïve Bayes and k-nearest neighbors models showed also valuable results (AUC ≈ 0.8) with a lower number of features (<13), thus suggesting that these models may be more generalizable when using external validations.

CONCLUSION

The proposed radiomics MRI approach is able to discriminate between GBM and BM with high accuracy employing a set of 2D texture features, thus helping in the diagnosis of brain lesions in a fast and non-invasive way.

Advanced Imaging of Brain Metastases: From Augmenting Visualization and Improving Diagnosis to Evaluating Treatment Response

Early detection of brain metastases and differentiation from other neuropathologies is crucial. Although biopsy is often required for definitive diagnosis, imaging can provide useful information. After treatment commences, imaging is also performed to assess the efficacy of treatment. Contrast-enhanced magnetic resonance imaging (MRI) is the traditional imaging method for the evaluation of brain metastases, as it provides information about lesion size, morphology, and macroscopic properties. Newer MRI sequences have been developed to increase the conspicuity of detecting enhancing metastases. Other advanced MRI techniques, that have the capability to probe beyond the anatomic structure, are available to characterize micro-structures, cellularity, physiology, perfusion, and metabolism. Artificial intelligence provides powerful computational tools for detection, segmentation, classification, prediction, and prognosis. We highlight and review a few advanced MRI techniques for the assessment of brain metastases-specifically for (1) diagnosis, including differentiating between malignancy types and (2) evaluation of treatment response, including the differentiation between radiation necrosis and disease progression.

Discrimination between Glioblastoma and Solitary Brain Metastasis: Comparison of Inflow-Based Vascular-Space-Occupancy and Dynamic Susceptibility Contrast MR Imaging

BACKGROUND AND PURPOSE

Accurate differentiation between glioblastoma and solitary brain metastasis is of vital importance clinically. This study aimed to investigate the potential value of the inflow-based vascular-space-occupancy MR imaging technique, which has no need for an exogenous contrast agent, in differentiating glioblastoma and solitary brain metastasis and to compare it with DSC MR imaging.

MATERIALS AND METHODS

Twenty patients with glioblastoma and 22 patients with solitary brain metastasis underwent inflow-based vascular-space-occupancy and DSC MR imaging with a 3T clinical scanner. Two neuroradiologists independently measured the maximum inflow-based vascular-space-occupancy-derived arteriolar CBV and DSC-derived CBV values in intratumoral regions and peritumoral T2-hyperintense regions, which were normalized to the contralateral white matter (relative arteriolar CBV and relative CBV, inflow-based vascular-space-occupancy relative arteriolar CBV, and DSC-relative CBV). The intraclass correlation coefficient, Student t test, or Mann-Whitney U test and receiver operating characteristic analysis were performed.

RESULTS

All parameters of both regions had good or excellent interobserver reliability (0.74∼0.89). In peritumoral T2-hyperintese regions, DSC-relative CBV (P < .001), inflow-based vascular-space-occupancy arteriolar CBV (P = .001), and relative arteriolar CBV (P = .005) were significantly higher in glioblastoma than in solitary brain metastasis, with areas under the curve of 0.94, 0.83, and 0.72 for discrimination, respectively. In the intratumoral region, both inflow-based vascular-space-occupancy arteriolar CBV and relative arteriolar CBV were significantly higher in glioblastoma than in solitary brain metastasis (both P < .001), with areas under the curve of 0.91 and 0.90, respectively. Intratumoral DSC-relative CBV showed no significant difference (P = .616) between the 2 groups.

CONCLUSIONS

Inflow-based vascular-space-occupancy has the potential to discriminate glioblastoma from solitary brain metastasis, especially in the intratumoral region.

Using mathematics in MRI data management for glioma assesment

Our aim is to review the mathematical tools usefulness in MR data management for glioma diagnosis and treatment optimization. MRI does not give access to organs variations in hours or days. However a lot of multiparametric data are generated. Mathematics could help to override this paradox, the aim of this article is to show how. We first make a review on mathematical modelling using equations. Afterwards we present statistical analysis. We provide detailed examples in both sections. We finally conclude, giving some clues on in silico models.

Differentiation of Recurrence from Radiation Necrosis in Gliomas Based on the Radiomics of Combinational Features and Multimodality MRI Images

Purpose

To classify radiation necrosis versus recurrence in glioma patients using a radiomics model based on combinational features and multimodality MRI images.

Methods

Fifty-one glioma patients who underwent radiation treatments after surgery were enrolled in this study. Sixteen patients revealed radiation necrosis while 35 patients showed tumor recurrence during the follow-up period. After treatment, all patients underwent T1-weighted, T1-weighted postcontrast, T2-weighted, and fluid-attenuated inversion recovery scans. A total of 41,284 handcrafted and 24,576 deep features were extracted for each patient. The 0.623 + bootstrap method and the area under the curve (denoted as 0.632 + bootstrap AUC) metric were used to select the features. The stepwise forward method was applied to construct 10 logistic regression models based on different combinations of image features.

Results

For handcrafted features on multimodality MRI, model 7 with seven features yielded the highest AUC of 0.9624, sensitivity of 0.8497, and specificity of 0.9083 in the validation set. These values were higher than the accuracy of using handcrafted features on single-modality MRI (paired t-test, p < 0.05, except sensitivity). For combined handcrafted and AlexNet features on multimodality MRI, model 6 with six features achieved the highest AUC of 0.9982, sensitivity of 0.9941, and specificity of 0.9755 in the validation set. These values were higher than the accuracy of using handcrafted features on multimodality MRI (paired t-test, p < 0.05).

Conclusions

Handcrafted and deep features extracted from multimodality MRI images reflecting the heterogeneity of gliomas can provide useful information for glioma necrosis/recurrence classification.

Discrimination Between Solitary Brain Metastasis and Glioblastoma Multiforme by Using ADC-Based Texture Analysis: A Comparison of Two Different ROI Placements

RATIONALE AND OBJECTIVES

To explore the value of texture analysis based on the apparent diffusion coefficient (ADC) value and the effect of region of interest (ROI) placements in distinguishing glioblastoma multiforme (GBM) from solitary brain metastasis (sMET).

MATERIALS AND METHODS

Sixty-two patients with pathologically confirmed GBM (n = 36) and sMET (n = 26) were retrospectively included. All patients underwent diffusion-weighted imaging with b values of 0 and 1000 s/mm², and the ADC maps were generated automatically. ROIs were placed on the largest whole single-slice tumor (ROI1) and the enhanced solid portion (ROI2) of the ADC maps, respectively. The texture feature metrics of the histogram and gray-level co-occurrence matrix were then extracted by using in-house software. The parameters of the texture analysis were compared between GBM and sMET, using the Mann-Whitney U test. A receiver operating characteristic (ROC) curve analysis was performed to determine the best parameters for distinguishing between GBM from sMET.

RESULTS

Homogeneity and the inverse difference moment (IDM) of GBM were significantly higher than those of sMET in both ROIs (ROI1, p = 0.014 for homogeneity and p = 0.048 for IDM; ROI2, p< 0.001 for homogeneity and p = 0.029 for IDM). According to the ROC curve analysis, the area under the ROC curve (AUC) of homogeneity in ROI1 (AUC, 0.682, sensitivity, 72.2%, specificity, 61.5%) was significantly lower than that of ROI2 (AUC, 0.886, sensitivity, 83.3%, specificity, 76.9%; p= 0.012), whereas the IDM showed no statistical significance between two ROIs (p> 0.05).

CONCLUSION

The ADC-based texture analysis can help differentiate GBM from sMET, and the ROI on the solid portion would be recommended to calculate the ADC-based texture metrics.

The Diagnostic Value of MRI-Based Texture Analysis in Discrimination of Tumors Located in Posterior Fossa: A Preliminary Study

Objectives

To investigate the diagnostic value of MRI-based texture analysis in discriminating common posterior fossa tumors, including medulloblastoma, brain metastatic tumor, and hemangioblastoma.

Methods

A total number of 185 patients were enrolled in the current study: 63 of them were diagnosed with medulloblastoma, 56 were diagnosed with brain metastatic tumor, and 66 were diagnosed with hemangioblastoma. Texture features were extracted from contrast-enhanced T1-weighted (T1C) images and fluid-attenuation inversion recovery (FLAIR) images within two matrixes. Mann–Whitney U test was conducted to identify whether texture features were significantly different among subtypes of tumors. Logistic regression analysis was performed to assess if they could be taken as independent predictors and to establish the integrated models. Receiver operating characteristic analysis was conducted to evaluate their performances in discrimination.

Results

There were texture features from both T1C images and FLAIR images found to be significantly different among the three types of tumors. The integrated model represented that the promising diagnostic performance of texture analysis depended on a series of features rather than a single feature. Moreover, the predictive model that combined texture features and clinical feature implied feasible performance in prediction with an accuracy of 0.80.

Conclusion

MRI-based texture analysis could potentially be served as a radiological method in discrimination of common tumors located in posterior fossa.

Machine Learning to Differentiate T2-Weighted Hyperintense Uterine Leiomyomas from Uterine Sarcomas by Utilizing Multiparametric Magnetic Resonance Quantitative Imaging Features

RATIONALE AND OBJECTIVE

Uterine leiomyomas with high signal intensity on T2-weighted imaging (T2WI) can be difficult to distinguish from sarcomas. This study assessed the feasibility of using machine learning to differentiate uterine sarcomas from leiomyomas with high signal intensity on T2WI on multiparametric magnetic resonance imaging.

MATERIALS AND METHODS

This retrospective study included 80 patients (50 with benign leiomyoma and 30 with uterine sarcoma) who underwent pelvic 3 T magnetic resonance imaging examination for the evaluation of uterine myometrial smooth muscle masses with high signal intensity on T2WI. We used six machine learning techniques to develop prediction models based on 12 texture parameters on T1WI and T2WI, apparent diffusion coefficient maps, and contrast-enhanced T1WI, as well as tumor size and age. We calculated the areas under the curve (AUCs) using receiver-operating characteristic analysis for each model by 10-fold cross-validation and compared these to those for two board-certified radiologists.

RESULTS

The eXtreme Gradient Boosting model gave the highest AUC (0.93), followed by the random forest, support vector machine, multilayer perceptron, k-nearest neighbors, and logistic regression models. Age was the most important factor for differentiation (leiomyoma 44.9 ± 11.1 years; sarcoma 58.9 ± 14.7 years; p < 0.001). The AUC for the eXtreme Gradient Boosting was significantly higher than those for both radiologists (0.93 vs 0.80 and 0.68, p = 0.03 and p < 0.001, respectively).

CONCLUSION

Machine learning outperformed experienced radiologists in the differentiation of uterine sarcomas from leiomyomas with high signal intensity on T2WI.

Glioblastoma and Anaplastic Astrocytoma: Differentiation Using MRI Texture Analysis

Introduction: Glioblastoma and anaplastic astrocytoma (ANA) are two of the most common primary brain tumors in adults. The differential diagnosis is important for treatment recommendations and prognosis assessment. This study aimed to assess the discriminative ability of texture analysis using machine learning to distinguish glioblastoma from ANA.

Methods: A total of 123 patients with glioblastoma (n = 76) or ANA (n = 47) were enrolled in this study. Texture features were extracted from contrast-enhanced Magnetic Resonance (MR) images using LifeX package. Three independent feature-selection methods were performed to select the most discriminating parameters:Distance Correlation, least absolute shrinkage and selection operator (LASSO), and gradient correlation decision tree (GBDT). These selected features (datasets) were then randomly split into the training and the validation group at the ratio of 4:1 and were fed into linear discriminant analysis (LDA), respectively, and independently. The standard sensitivity, specificity, the areas under receiver operating characteristic curve (AUC) and accuracy were calculated for both training and validation group.

Results: All three models (Distance Correlation + LDA, LASSO + LDA and GBDT + LDA) showed promising ability to discriminate glioblastoma from ANA, with AUCs ≥0.95 for both the training and the validation group using LDA algorithm and no overfitting was observed. LASSO + LDA showed the best discriminative ability in horizontal comparison among three models.

Conclusion: Our study shows that MRI texture analysis using LDA algorithm had promising ability to discriminate glioblastoma from ANA. Multi-center studies with greater number of patients are warranted in future studies to confirm the preliminary result.

Utility of Percentage Signal Recovery and Baseline Signal in DSC-MRI Optimized for Relative CBV Measurement for Differentiating Glioblastoma, Lymphoma, Metastasis, and Meningioma

BACKGROUND AND PURPOSE

The percentage signal recovery in non-leakage-corrected (no preload, high flip angle, intermediate TE) DSC-MR imaging is known to differ significantly for glioblastoma, metastasis, and primary CNS lymphoma. Because the percentage signal recovery is influenced by preload and pulse sequence parameters, we investigated whether the percentage signal recovery can still differentiate these common contrast-enhancing neoplasms using a DSC-MR imaging protocol designed for relative CBV accuracy (preload, intermediate flip angle, low TE).

MATERIALS AND METHODS

We retrospectively analyzed DSC-MR imaging of treatment-naïve, pathology-proved glioblastomas (n = 14), primary central nervous system lymphomas (n = 7), metastases (n = 20), and meningiomas (n = 13) using a protocol designed for relative CBV accuracy (a one-quarter-dose preload and single-dose bolus of gadobutrol, TR/TE = 1290/40 ms, flip angle = 60° at 1.5T). Mean percentage signal recovery, relative CBV, and normalized baseline signal intensity were compared within contrast-enhancing lesion volumes. Classification accuracy was determined by receiver operating characteristic analysis.

RESULTS

Relative CBV best differentiated meningioma from glioblastoma and from metastasis with areas under the curve of 0.84 and 0.82, respectively. The percentage signal recovery best differentiated primary central nervous system lymphoma from metastasis with an area under the curve of 0.81. Relative CBV and percentage signal recovery were similar in differentiating primary central nervous system lymphoma from glioblastoma and from meningioma. Although neither relative CBV nor percentage signal recovery differentiated glioblastoma from metastasis, mean normalized baseline signal intensity achieved 86% sensitivity and 50% specificity.

CONCLUSIONS

Similar to results for non-preload-based DSC-MR imaging, percentage signal recovery for one-quarter-dose preload-based, intermediate flip angle DSC-MR imaging differentiates most pair-wise comparisons of glioblastoma, metastasis, primary central nervous system lymphoma, and meningioma, except for glioblastoma versus metastasis. Differences in normalized post-preload baseline signal for glioblastoma and metastasis, reflecting a snapshot of dynamic contrast enhancement, may motivate the use of single-dose multiecho protocols permitting simultaneous quantification of DSC-MR imaging and dynamic contrast-enhanced MR imaging parameters.

Computational quantitative MR image features - a potential useful tool in differentiating glioblastoma from solitary brain metastasis

PURPOSE

Glioblastomas (GBM) and metastases are the most frequent malignant brain tumors in the adult population. Their presentation on conventional MRI is quite similar, but treatment strategy and prognosis are substantially different. Even with advanced MR techniques, in some cases diagnostic uncertainty remains. The main objective of this study was to determine whether fractal, texture, or both MR image analyses could aid in differentiating glioblastoma from solitary brain metastasis.

METHOD

In a retrospective study of 55 patients (30 glioblastomas and 25 solitary metastases) who underwent T2W/SWI/CET1 MRI, quantitative parameters of fractal and texture analysis were estimated, using box-counting and gray level co-occurrence matrix (GLCM) methods.

RESULTS

All five GLCM parameters obtained from T2W images showed significant difference between glioblastomas and solitary metastases, as well as on CET1 images except correlation (S_COR), contrary to SWI images which showed different values of two parameters (angular second moment-S_ASM and contrast-S_CON). Only three fractal features (binary box dimension-D_bin, normalized box dimension-D_norm and lacunarity-λ) measured on T2W and D_norm measured on CET1 images significantly differed GBMs from solitary metastases. The highest sensitivity and specificity were obtained from inverse difference moment (S_IDM) on T2W and S_IDM on CET1 images, respectively. Combination of several GLCM parameters yielded better results. The processing of T2W images provided the most significantly different parameters between the groups, followed by CET1 and SWI images.

CONCLUSIONS

Computational-aided quantitative image analysis may potentially improve diagnostic accuracy. According to our results texture features are more significant than fractal-based features in differentiation glioblastoma from solitary metastasis.

Texture analysis on diffusion tensor imaging: discriminating glioblastoma from single brain metastasis

BACKGROUND

Texture analysis has been done on several radiological modalities to stage, differentiate, and predict prognosis in many oncologic tumors.

PURPOSE

To determine the diagnostic accuracy of discriminating glioblastoma (GBM) from single brain metastasis (MET) by assessing the heterogeneity of both the solid tumor and the peritumoral edema with magnetic resonance imaging (MRI) texture analysis (MRTA).

MATERIAL AND METHODS

Preoperative MRI examinations done on a 3-T scanner of 43 patients were included: 22 GBM and 21 MET. MRTA was performed on diffusion tensor imaging (DTI) in a representative region of interest (ROI). The MRTA was assessed using a commercially available research software program (TexRAD) which applies a filtration histogram technique for characterizing tumor and peritumoral heterogeneity. The filtration step selectively filters and extracts texture features at different anatomical scales varying from 2 mm (fine) to 6 mm (coarse). Heterogeneity quantification was obtained by the statistical parameter entropy. A threshold value to differentiate GBM from MET with sensitivity and specificity was calculated by receiver operating characteristic (ROC) analysis.

RESULTS

Quantifying the heterogeneity of the solid part of the tumor showed no significant difference between GBM and MET. However, the heterogeneity of the GBMs peritumoral edema was significantly higher than the edema surrounding MET, differentiating them with a sensitivity of 80% and specificity of 90%.

CONCLUSION

Assessing the peritumoral heterogeneity can increase the radiological diagnostic accuracy when discriminating GBM and MET. This will facilitate the medical staging and optimize the planning for surgical resection of the tumor and postoperative management.

Machine-learning in grading of gliomas based on multi-parametric magnetic resonance imaging at 3T

OBJECTIVE

The objective of this study was to assess the contribution of multi-parametric (mp) magnetic resonance imaging (MRI) quantitative features in the machine learning-based grading of gliomas with a multi-region-of-interests approach.

MATERIALS AND METHODS

Forty-three patients who were newly diagnosed as having a glioma were included in this study. The patients were scanned prior to any therapy using a standard brain tumor magnetic resonance (MR) imaging protocol that included T1 and T2-weighted, diffusion-weighted, diffusion tensor, MR perfusion and MR spectroscopic imaging. Three different regions-of-interest were drawn for each subject to encompass tumor, immediate tumor periphery, and distant peritumoral edema/normal. The normalized mp-MRI features were used to build machine-learning models for differentiating low-grade gliomas (WHO grades I and II) from high grades (WHO grades III and IV). In order to assess the contribution of regional mp-MRI quantitative features to the classification models, a support vector machine-based recursive feature elimination method was applied prior to classification.

RESULTS

A machine-learning model based on support vector machine algorithm with linear kernel achieved an accuracy of 93.0%, a specificity of 86.7%, and a sensitivity of 96.4% for the grading of gliomas using ten-fold cross validation based on the proposed subset of the mp-MRI features.

CONCLUSION

In this study, machine-learning based on multiregional and multi-parametric MRI data has proven to be an important tool in grading glial tumors accurately even in this limited patient population. Future studies are needed to investigate the use of machine learning algorithms for brain tumor classification in a larger patient cohort.

Prediction of successful shock wave lithotripsy with CT: a phantom study using texture analysis

OBJECTIVE

To apply texture analysis (TA) in computed tomography (CT) of urinary stones and to correlate TA findings with the number of required shockwaves for successful shock wave lithotripsy (SWL).

MATERIALS AND METHODS

CT was performed on thirty-four urinary stones in an in vitro setting. Urinary stones underwent SWL and the number of required shockwaves for disintegration was recorded. TA was performed after post-processing for pixel spacing and image normalization. Feature selection and dimension reduction were performed according to inter- and intrareader reproducibility and by evaluating the predictive ability of the number of shock waves with the degree of redundancy between TA features. Three regression models were tested: (1) linear regression with elimination of colinear attributes (2), sequential minimal optimization regression (SMOreg) employing machine learning, and (3) simple linear regression model of a single TA feature with lowest squared error.

RESULTS

Highest correlations with the absolute number of required SWL shockwaves were found for the linear regression model (r = 0.55, p = 0.005) using two weighted TA features: Histogram 10th Percentile, and Gray-Level Co-Occurrence Matrix (GLCM) S(3, 3) SumAverg. Using the median number of required shockwaves (n = 72) as a threshold, receiver-operating characteristic analysis showed largest area-under-the-curve values for the SMOreg model (AUC = 0.84, r = 0.51, p < 0.001) using four weighted TA features: Histogram 10^th Percentile, and GLCM S(1, 1) InvDfMom, S(3, 3) SumAverg, and S(4, -4) SumVarnc.

CONCLUSION

Our in vitro study illustrates the proof-of-principle of TA of urinary stone CT images for predicting the success of stone disintegration with SWL.

Perfusion MRI as a diagnostic biomarker for differentiating glioma from brain metastasis: a systematic review and meta-analysis

OBJECTIVES

Differentiation of glioma from brain metastasis is clinically crucial because it affects the clinical outcome of patients and alters patient management. Here, we present a systematic review and meta-analysis of the currently available data on perfusion magnetic resonance imaging (MRI) for differentiating glioma from brain metastasis, assessing MRI protocols and parameters.

METHODS

A computerised search of Ovid-MEDLINE and EMBASE databases was performed up to 3 October 2017, to find studies on the diagnostic performance of perfusion MRI for differentiating glioma from brain metastasis. Pooled summary estimates of sensitivity and specificity were obtained using hierarchical logistic regression modelling. We conducted meta-regression and subgroup analyses to explain the effects of the study heterogeneity.

RESULTS

Eighteen studies with 900 patients were included. The pooled sensitivity and specificity were 90% (95% CI, 84-94%) and 91% (95% CI, 84-95%), respectively. The area under the hierarchical summary receiver operating characteristic curve was 0.96 (95% CI, 0.94-0.98). The meta-regression showed that the percentage of glioma in the study population and the study design were significant factors affecting study heterogeneity. In a subgroup analysis including patients with glioblastoma only, the pooled sensitivity was 92% (95% CI, 84-97%) and the pooled specificity was 94% (95% CI, 85-98%).

CONCLUSIONS

Although various perfusion MRI techniques were used, the current evidence supports the use of perfusion MRI to differentiate glioma from brain metastasis. In particular, perfusion MRI showed excellent diagnostic performance for differentiating glioblastoma from brain metastasis.

KEY POINTS

• Perfusion MRI shows high diagnostic performance for differentiating glioma from brain metastasis. • The pooled sensitivity was 90% and pooled specificity was 91%. • Peritumoral rCBV derived from DSC is a relatively well-validated.

Applying protein-based amide proton transfer MR imaging to distinguish solitary brain metastases from glioblastoma

OBJECTIVES

To determine the utility of amide proton transfer-weighted (APTw) MR imaging in distinguishing solitary brain metastases (SBMs) from glioblastomas (GBMs).

METHODS

Forty-five patients with SBMs and 43 patients with GBMs underwent conventional and APT-weighted sequences before clinical intervention. The APTw parameters and relative APTw (rAPTw) parameters in the tumour core and the peritumoral brain zone (PBZ) were obtained and compared between SBMs and GBMs. The receiver-operating characteristic (ROC) curve was used to assess the best parameter for distinguishing between the two groups.

RESULTS

The APTw_max, APTw_min, APTw_mean, rAPTw_max, rAPTw_min or rAPTw_mean values in the tumour core were not significantly different between the SBM and GBM groups (P = 0.141, 0.361, 0.221, 0.305, 0.578 and 0.448, respectively). However, the APTw_max, APTw_min, APTw_mean, rAPTw_max, rAPTw_min or rAPTw_mean values in the PBZ were significantly lower in the SBM group than in the GBM group (P < 0.001). The APTw_min values had the highest area under the ROC curve 0.905 and accuracy 85.2% in discriminating between the two neoplasms.

CONCLUSION

As a noninvasive imaging method, APT-weighted MR imaging can be used to distinguish SBMs from GBMs.

KEY POINTS

• APTw values in the tumour core were not different between SBMs and GBMs. • APTw values in peritumoral brain zone were lower in SBMs than in GBMs. • The APTw _min was the best parameter to distinguish SBMs from GBMs.

Differentiation of Recurrent Glioblastoma from Delayed Radiation Necrosis by Using Voxel-based Multiparametric Analysis of MR Imaging Data

Purpose To assess a volume-weighted voxel-based multiparametric (MP) clustering method as an imaging biomarker to differentiate recurrent glioblastoma from delayed radiation necrosis. Materials and Methods The institutional review board approved this retrospective study and waived the informed consent requirement. Seventy-five patients with pathologic analysis-confirmed recurrent glioblastoma (n = 42) or radiation necrosis (n = 33) who presented with enlarged contrast material-enhanced lesions at magnetic resonance (MR) imaging after they completed concurrent chemotherapy and radiation therapy were enrolled. The diagnostic performance of the total MP cluster score was determined by using the area under the receiver operating characteristic curve (AUC) with cross-validation and compared with those of single parameter measurements (10% histogram cutoffs of apparent diffusion coefficient [ADC10] or 90% histogram cutoffs of normalized cerebral blood volume and initial time-signal intensity AUC). Results Receiver operating characteristic curve analysis showed that an AUC for differentiating recurrent glioblastoma from delayed radiation necrosis was highest in the total MP cluster score and lowest for ADC10 for both readers. The total MP cluster score had significantly better diagnostic accuracy than any single parameter (corrected P = .001-.039 for reader 1; corrected P = .005-.041 for reader 2). The total MP cluster score was the best predictor of recurrent glioblastoma (cross-validated AUCs, 0.942-0.946 for both readers), with a sensitivity of 95.2% for reader 1 and 97.6% for reader 2. Conclusion Quantitative analysis with volume-weighted voxel-based MP clustering appears to be superior to the use of single imaging parameters to differentiate recurrent glioblastoma from delayed radiation necrosis. ^© RSNA, 2017 Online supplemental material is available for this article.

Radiomic Phenotyping in Brain Cancer to Unravel Hidden Information in Medical Images

Radiomics is a new area of research in the field of imaging with tremendous potential to unravel the hidden information in digital images. The scope of radiology has grown exponentially over the last two decades; since the advent of radiomics, many quantitative imaging features can now be extracted from medical images through high-throughput computing, and these can be converted into mineable data that can help in linking imaging phenotypes with clinical data, genomics, proteomics, and other "omics" information. In cancer, radiomic imaging analysis aims at extracting imaging features embedded in the imaging data, which can act as a guide in the disease or cancer diagnosis, staging and planning interventions for treating patients, monitor patients on therapy, predict treatment response, and determine patient outcomes.

Comparison of Cerebral Blood Volume and Plasma Volume in Untreated Intracranial Tumors

Purpose

Plasma volume and blood volume are imaging-derived parameters that are often used to evaluation intracranial tumors. Physiologically, these parameters are directly related, but their two different methods of measurements, T1-dynamic contrast enhanced (DCE)- and T2-dynamic susceptibility contrast (DSC)-MR utilize different model assumptions and approaches. This poses the question of whether the interchangeable use of T1-DCE-MRI derived fractionated plasma volume (vp) and relative cerebral blood volume (rCBV) assessed using DSC-MRI, particularly in glioblastoma, is reliable, and if this relationship can be generalized to other types of brain tumors. Our goal was to examine the hypothetical correlation between these parameters in three most common intracranial tumor types.

Methods

Twenty-four newly diagnosed, treatment naïve brain tumor patients, who had undergone DCE- and DSC-MRI, were classified in three histologically proven groups: glioblastoma (n = 7), meningioma (n = 9), and intraparenchymal metastases (n = 8). The rCBV was obtained from DSC after normalization with the normal-appearing anatomically symmetrical contralateral white matter. Correlations between these parameters were evaluated using Pearson (r), Spearman's (ρ) and Kendall’s tau-b (τ_B) rank correlation coefficient.

Results

The Pearson, Spearman and Kendall’s correlation between vp with rCBV were r = 0.193, ρ = 0.253 and τ_B = 0.33 (p-_Pearson = 0.326, p-_Spearman= 0.814 and p-_Kendall= 0.823) in glioblastoma, r = -0.007, ρ = 0.051 and τ_B = 0.135 (p-_Pearson = 0.970, p-_Spearman= 0.765 and p-_Kendall= 0.358) in meningiomas, and r = 0.289, ρ = 0.228 and τ_B = 0.239 (p-_Pearson = 0.109, p-_Spearman= 0.210 and p-_Kendall= 0.095) in metastasis.

Conclusion

Results indicate that no correlation exists between vp with rCBV in glioblastomas, meningiomas and intraparenchymal metastatic lesions. Consequently, these parameters, as calculated in this study, should not be used interchangeably in either research or clinical practice.

Quantitative Perfusion and Permeability Biomarkers in Brain Cancer from Tomographic CT and MR Images

Dynamic contrast-enhanced perfusion and permeability imaging, using computed tomography and magnetic resonance systems, are important techniques for assessing the vascular supply and hemodynamics of healthy brain parenchyma and tumors. These techniques can measure blood flow, blood volume, and blood-brain barrier permeability surface area product and, thus, may provide information complementary to clinical and pathological assessments. These have been used as biomarkers to enhance the treatment planning process, to optimize treatment decision-making, and to enable monitoring of the treatment noninvasively. In this review, the principles of magnetic resonance and computed tomography dynamic contrast-enhanced perfusion and permeability imaging are described (with an emphasis on their commonalities), and the potential values of these techniques for differentiating high-grade gliomas from other brain lesions, distinguishing true progression from posttreatment effects, and predicting survival after radiotherapy, chemotherapy, and antiangiogenic treatments are presented.

Accuracy of percentage of signal intensity recovery and relative cerebral blood volume derived from dynamic susceptibility-weighted, contrast-enhanced MRI in the preoperative diagnosis of cerebral tumours

Conventional magnetic resonance imaging (MRI) is the technique of choice for diagnosis of cerebral tumours, and has become an increasingly powerful tool for their evaluation; however, the diagnosis of common contrast-enhancing lesions can be challenging, as it is sometimes impossible to differentiate them using conventional imaging. Histopathological analysis of biopsy specimens is the gold standard for diagnosis; however, there are significant risks associated with the invasive procedure and definitive diagnosis is not always achieved. Early accurate diagnosis is important, as management differs accordingly. Advanced MRI techniques have increasing utility for aiding diagnosis in a variety of clinical scenarios. Dynamic susceptibility-weighted contrast-enhanced (DSC) MRI is a perfusion imaging technique and a potentially important tool for the characterisation of cerebral tumours. The percentage of signal intensity recovery (PSR) and relative cerebral blood volume (rCBV) derived from DSC MRI provide information about tumour capillary permeability and neoangiogenesis, which can be used to characterise tumour type and grade, and distinguish tumour recurrence from treatment-related effects. Therefore, PSR and rCBV potentially represent a non-invasive means of diagnosis; however, the clinical utility of these parameters has yet to be established. We present a review of the literature to date.

Quantification of heterogeneity as a biomarker in tumor imaging: a systematic review

BACKGROUND

Many techniques are proposed for the quantification of tumor heterogeneity as an imaging biomarker for differentiation between tumor types, tumor grading, response monitoring and outcome prediction. However, in clinical practice these methods are barely used. This study evaluates the reported performance of the described methods and identifies barriers to their implementation in clinical practice.

METHODOLOGY

The Ovid, Embase, and Cochrane Central databases were searched up to 20 September 2013. Heterogeneity analysis methods were classified into four categories, i.e., non-spatial methods (NSM), spatial grey level methods (SGLM), fractal analysis (FA) methods, and filters and transforms (F&T). The performance of the different methods was compared.

PRINCIPAL FINDINGS

Of the 7351 potentially relevant publications, 209 were included. Of these studies, 58% reported the use of NSM, 49% SGLM, 10% FA, and 28% F&T. Differentiation between tumor types, tumor grading and/or outcome prediction was the goal in 87% of the studies. Overall, the reported area under the curve (AUC) ranged from 0.5 to 1 (median 0.87). No relation was found between the performance and the quantification methods used, or between the performance and the imaging modality. A negative correlation was found between the tumor-feature ratio and the AUC, which is presumably caused by overfitting in small datasets. Cross-validation was reported in 63% of the classification studies. Retrospective analyses were conducted in 57% of the studies without a clear description.

CONCLUSIONS

In a research setting, heterogeneity quantification methods can differentiate between tumor types, grade tumors, and predict outcome and monitor treatment effects. To translate these methods to clinical practice, more prospective studies are required that use external datasets for validation: these datasets should be made available to the community to facilitate the development of new and improved methods.

Microvascular MRI and unsupervised clustering yields histology-resembling images in two rat models of glioma

Imaging heterogeneous cancer lesions is a real challenge. For diagnosis, histology often remains the reference, but it is widely acknowledged that biopsies are not reliable. There is thus a strong interest in establishing a link between clinical in vivo imaging and the biologic properties of tissues. In this study, we propose to construct histology-resembling images based on tissue microvascularization, a magnetic resonance imaging (MRI) accessible source of contrast. To integrate the large amount of information collected with microvascular MRI, we combined a manual delineation of a spatial region of interest with an unsupervised, model-based cluster analysis (Mclust). This approach was applied to two rat models of glioma (C6 and F98). Six MRI parameters were mapped: apparent diffusion coefficient, vessel wall permeability, cerebral blood volume fraction, cerebral blood flow, tissular oxygen saturation, and cerebral metabolic rate of oxygen. Five clusters, defined by their MRI features, were found to correspond to specific histologic features, and revealed intratumoral spatial structures. These results suggest that the presence of a cluster within a tumor can be used to assess the presence of a tissue type. In addition, the cluster composition, i.e., a signature of the intratumoral structure, could be used to characterize tumor models as histology does.

The effect of pulse sequence parameters and contrast agent dose on percentage signal recovery in DSC-MRI: implications for clinical applications

BACKGROUND AND PURPOSE

Both technical and pathophysiologic factors affect PSR in DSC-MR imaging. We aimed to determine how TE, flip angle (α), and contrast dose impact PSR in high-grade gliomas.

MATERIALS AND METHODS

We retrospectively computed PSR maps for 22 patients with high-grade gliomas, comparing 3 DSC-MR imaging methods by using single-dose gadodiamide without preload administration: A (n = 7), α = 35°, TE = 54 ms; B (n = 5), α = 72°, TE = 30 ms; C (n = 10), α = 90°, TE = 30 ms. Methods A-C served as preload for subsequent dynamic imaging using method D (method C parameters but with double-dose contrast). We compared first- and second-injection tumor PSR for methods C and D (paired t test) and tumor PSR for both injections grouped by the first-injection acquisition method (3-group nonparametric 1-way ANOVA). We compared PSR in tumor and normal brain for each first- and second-injection method group (paired t test).

RESULTS

First-injection PSR in tumor and normal brain differed significantly for methods B (P = .01) and C (P = .05), but not A (P = .71). First-injection tumor PSR increased with T1 weighting with a significant main effect of method groupings (P = .0012), but there was no significant main effect for first-injection normal brain (P = .93), or second-injection tumor (P = .95) or normal brain (P = .13). In patients scanned with methods C and D, first-injection PSR significantly exceeded second-injection PSR for tumor (P = .037) and normal brain (P < .001).

CONCLUSIONS

PSR strongly depends on the T1 weighting of DSC-MR imaging, including pulse sequence (TE, α) and contrast agent (dose, preload) parameters, with implications for protocol design and the interpretation and comparison of PSR values across tumor types and imaging centers.

Pre- and post-contrast three-dimensional double inversion-recovery MRI in human glioblastoma

Fluid attenuated inversion recovery (FLAIR) MRI sequences have become an indispensible tool for defining the malignant boundary in patients with brain tumors by nulling the signal contribution from cerebrospinal fluid allowing both regions of edema and regions of non-enhancing, infiltrating tumor to become hyperintense on resulting images. In the current study we examined the utility of a three-dimensional double inversion recovery (DIR) sequence that additionally nulls the MR signal associated with white matter, implemented either pre-contrast or post-contrast, in order to determine whether this sequence allows for better differentiation between tumor and normal brain tissue. T1- and T2-weighted, FLAIR, dynamic susceptibility contrast (DSC)-MRI estimates of cerebral blood volume (rCBV), contrast-enhanced T1-weighted images (T1+C), and DIR data (pre- or post-contrast) were acquired in 22 patients with glioblastoma. Contrast-to-noise (CNR) and tumor volumes were compared between DIR and FLAIR sequences. Line profiles across regions of tumor were generated to evaluate similarities between image contrasts. Additionally, voxel-wise associations between DIR and other sequences were examined. Results suggested post-contrast DIR images were hyperintense (bright) in regions spatially similar those having FLAIR hyperintensity and hypointense (dark) in regions with contrast-enhancement or elevated rCBV due to the high sensitivity of 3D turbo spin echo sequences to susceptibility differences between different tissues. DIR tumor volumes were statistically smaller than tumor volumes as defined by FLAIR (Paired t test, P = 0.0084), averaging a difference of approximately 14 mL or 24 %. DIR images had approximately 1.5× higher lesion CNR compared with FLAIR images (Paired t test, P = 0.0048). Line profiles across tumor regions and scatter plots of voxel-wise coherence between different contrasts confirmed a positive correlation between DIR and FLAIR signal intensity and a negative correlation between DIR and both post-contrast T1-weighted image signal intensity and rCBV. Additional discrepancies between FLAIR and DIR abnormal regions were also observed, together suggesting DIR may provide additional information beyond that of FLAIR.