The role of diffusion and perfusion weighted imaging in the differential diagnosis of cerebral tumors: a review and future perspectives

Progression/Recurrence of Meningioma: An Imaging Review Based on Magnetic Resonance Imaging

Meningiomas are the most common primary central nervous system tumors. The preferred treatment is maximum safe resection, and the heterogeneity of meningiomas results in a variable prognosis. Progression/recurrence (P/R) can occur at any grade of meningioma and is a common adverse outcome after surgical treatment and a major cause of postoperative rehospitalization, secondary surgery, and mortality. Early prediction of P/R plays an important role in postoperative management, further adjuvant therapy, and follow-up of patients. Therefore, it is essential to thoroughly analyze the heterogeneity of meningiomas and predict postoperative P/R with the aid of noninvasive preoperative imaging. In recent years, the development of advanced magnetic resonance imaging technology and machine learning has provided new insights into noninvasive preoperative prediction of meningioma P/R, which helps to achieve accurate prediction of meningioma P/R. This narrative review summarizes the current research on conventional magnetic resonance imaging, functional magnetic resonance imaging, and machine learning in predicting meningioma P/R. We further explore the significance of tumor microenvironment in meningioma P/R, linking imaging features with tumor microenvironment to comprehensively reveal tumor heterogeneity and provide new ideas for future research.

Deep learning with diffusion MRI as in vivo microscope reveals sex-related differences in human white matter microstructure

Biological sex is a crucial variable in neuroscience studies where sex differences have been documented across cognitive functions and neuropsychiatric disorders. While gross statistical differences have been previously documented in macroscopic brain structure such as cortical thickness or region size, less is understood about sex-related cellular-level microstructural differences which could provide insight into brain health and disease. Studying these microstructural differences between men and women paves the way for understanding brain disorders and diseases that manifest differently in different sexes. Diffusion MRI is an important in vivo, non-invasive methodology that provides a window into brain tissue microstructure. Our study develops multiple end-to-end classification models that accurately estimates the sex of a subject using volumetric diffusion MRI data and uses these models to identify white matter regions that differ the most between men and women. 471 male and 560 female healthy subjects (age range, 22-37 years) from the Human Connectome Project are included. Fractional anisotropy, mean diffusivity and mean kurtosis are used to capture brain tissue microstructure characteristics. Diffusion parametric maps are registered to a standard template to reduce bias that can arise from macroscopic anatomical differences like brain size and contour. This study employ three major model architectures: 2D convolutional neural networks, 3D convolutional neural networks and Vision Transformer (with self-supervised pretraining). Our results show that all 3 models achieve high sex classification performance (test AUC 0.92-0.98) across all diffusion metrics indicating definitive differences in white matter tissue microstructure between males and females. We further use complementary model architectures to inform about the pattern of detected microstructural differences and the influence of short-range versus long-range interactions. Occlusion analysis together with Wilcoxon signed-rank test is used to determine which white matter regions contribute most to sex classification. The results indicate that sex-related differences manifest in both local features as well as global features / longer-distance interactions of tissue microstructure. Our highly consistent findings across models provides new insight supporting differences between male and female brain cellular-level tissue organization particularly in the central white matter.

High-performance presurgical differentiation of glioblastoma and metastasis by means of multiparametric neurite orientation dispersion and density imaging (NODDI) radiomics

OBJECTIVES

To evaluate the performance of multiparametric neurite orientation dispersion and density imaging (NODDI) radiomics in distinguishing between glioblastoma (Gb) and solitary brain metastasis (SBM).

MATERIALS AND METHODS

In this retrospective study, NODDI images were curated from 109 patients with Gb (n = 57) or SBM (n = 52). Automatically segmented multiple volumes of interest (VOIs) encompassed the main tumor regions, including necrosis, solid tumor, and peritumoral edema. Radiomics features were extracted for each main tumor region, using three NODDI parameter maps. Radiomics models were developed based on these three NODDI parameter maps and their amalgamation to differentiate between Gb and SBM. Additionally, radiomics models were constructed based on morphological magnetic resonance imaging (MRI) and diffusion imaging (diffusion-weighted imaging [DWI]; diffusion tensor imaging [DTI]) for performance comparison.

RESULTS

The validation dataset results revealed that the performance of a single NODDI parameter map model was inferior to that of the combined NODDI model. In the necrotic regions, the combined NODDI radiomics model exhibited less than ideal discriminative capabilities (area under the receiver operating characteristic curve [AUC] = 0.701). For peritumoral edema regions, the combined NODDI radiomics model achieved a moderate level of discrimination (AUC = 0.820). Within the solid tumor regions, the combined NODDI radiomics model demonstrated superior performance (AUC = 0.904), surpassing the models of other VOIs. The comparison results demonstrated that the NODDI model was better than the DWI and DTI models, while those of the morphological MRI and NODDI models were similar.

CONCLUSION

The NODDI radiomics model showed promising performance for preoperative discrimination between Gb and SBM.

CLINICAL RELEVANCE STATEMENT

The NODDI radiomics model showed promising performance for preoperative discrimination between Gb and SBM, and radiomics features can be incorporated into the multidimensional phenotypic features that describe tumor heterogeneity.

KEY POINTS

• The neurite orientation dispersion and density imaging (NODDI) radiomics model showed promising performance for preoperative discrimination between glioblastoma and solitary brain metastasis. • Compared with other tumor volumes of interest, the NODDI radiomics model based on solid tumor regions performed best in distinguishing the two types of tumors. • The performance of the single-parameter NODDI model was inferior to that of the combined-parameter NODDI model.

Brain Tumor Imaging: Review of Conventional and Advanced Techniques

Approaches to central nervous system (CNS) tumor classification and evaluation have undergone multiple iterations over the past few decades, in large part due to our growing understanding of the influence of genetics on tumor behavior and our refinement of brain tumor imaging techniques. Computed tomography and magnetic resonance imaging (MRI) both play a critical role in the diagnosis and monitoring of brain tumors, although MRI has become especially important due to its superior soft tissue resolution. The purpose of this article will be to briefly review the fundamentals of conventional and advanced techniques used in brain tumor imaging. We will also highlight the applications of these imaging tools in the context of commonly encountered tumors based on the most recently updated 2021 World Health Organization (WHO) classification of CNS tumors framework.

Comprehensive Brain Tumour Characterisation with VERDICT-MRI: Evaluation of Cellular and Vascular Measures Validated by Histology

The aim of this work was to extend the VERDICT-MRI framework for modelling brain tumours, enabling comprehensive characterisation of both intra- and peritumoural areas with a particular focus on cellular and vascular features. Diffusion MRI data were acquired with multiple b-values (ranging from 50 to 3500 s/mm2), diffusion times, and echo times in 21 patients with brain tumours of different types and with a wide range of cellular and vascular features. We fitted a selection of diffusion models that resulted from the combination of different types of intracellular, extracellular, and vascular compartments to the signal. We compared the models using criteria for parsimony while aiming at good characterisation of all of the key histological brain tumour components. Finally, we evaluated the parameters of the best-performing model in the differentiation of tumour histotypes, using ADC (Apparent Diffusion Coefficient) as a clinical standard reference, and compared them to histopathology and relevant perfusion MRI metrics. The best-performing model for VERDICT in brain tumours was a three-compartment model accounting for anisotropically hindered and isotropically restricted diffusion and isotropic pseudo-diffusion. VERDICT metrics were compatible with the histological appearance of low-grade gliomas and metastases and reflected differences found by histopathology between multiple biopsy samples within tumours. The comparison between histotypes showed that both the intracellular and vascular fractions tended to be higher in tumours with high cellularity (glioblastoma and metastasis), and quantitative analysis showed a trend toward higher values of the intracellular fraction (fic) within the tumour core with increasing glioma grade. We also observed a trend towards a higher free water fraction in vasogenic oedemas around metastases compared to infiltrative oedemas around glioblastomas and WHO 3 gliomas as well as the periphery of low-grade gliomas. In conclusion, we developed and evaluated a multi-compartment diffusion MRI model for brain tumours based on the VERDICT framework, which showed agreement between non-invasive microstructural estimates and histology and encouraging trends for the differentiation of tumour types and sub-regions.

Evaluation of the Value of Perfusion-Weighted Magnetic Resonance Imaging in the Differential Diagnosis of Sellar and Parasellar Tumors

The purpose of this study was to assess the value of perfusion-weighted imaging (PWI) in the differential diagnosis of sellar and parasellar tumors, as an additional sequence in the magnetic resonance imaging (MRI) protocol. Analysis was based on a substantial group of subjects and included 124 brain and pituitary MRI examinations with a dynamic susceptibility contrast (DSC) PWI sequence. The following perfusion parameters were determined for the tumors: relative cerebral blood volume (rCBV), relative peak height (rPH) and relative percentage of signal intensity recovery (rPSR). To ensure greater repeatability, each of the aforementioned parameters was calculated as: arithmetic mean of the values of the whole tumor, arithmetic mean of the maximum values on each axial slice within the tumor and maximum values derived from the whole tumor. In our study, we established that meningiomas compared to both non-functional and hormone-secreting pituitary adenomas (pituitary neuroendocrine tumors-PitNET) had significantly higher values of rCBV with cut-off points set at 3.45 and 3.54, respectively (mean rCBV). Additionally, meningiomas presented significantly higher maximum and mean maximum rPH values compared to adenomas. DSC PWI imaging adds significant value to conventional MRI examinations and can be helpful in differentiating equivocal pituitary tumors.

Value of quantitative apparent diffusion coefficients in differentiating low-grade gliomas from mixed neuronal-glial tumors

Purpose

To retrospectively assess if diffusion-weighted MR imaging (DWI) and quantitative apparent-diffusion coefficient (ADC) maps could be used to differentiate between low-grade gliomas (LGGs) and mixed neuronal-glial tumors (MNGTs including Dysembryoplastic Neuroepithelial Tumor and Ganglioglioma).

Materials and methods

We retrospectively searched the clinical, pathological, and radiological databases for a span of 9 years and identified 24 patients with biopsy proven LGG. This included WHO (fourth edition) grade I and II tumors including astrocytoma, oligoastrocytoma and oligodendrogliomas. We also identified 22 patients with MNGTs (WHO grade I) including 13 patients with DNET and 9 patients with Ganglioglioma. All patients with pathologically confirmed tumors who had MRI including DWI sequence were included in the study. Regions of interest (ROIs) of 0.1-0.15 cm² were manually positioned on the ADC maps and multiple values (10^-6 mm²/s) were obtained including the ADCmean. Optimal thresholds of ADC values and ADC ratios for distinguishing low-grade gliomas from mixed neuronal-glial tumors were determined by receiver operating characteristic (ROC) curve analysis.

Results

All the four ADC measurement variables, including the minimum (ADC min), the (ADC max) maximum, the mean of ADC values (ADC mean) and the ADC ratios (ADC mean/ADCnormal) showed significant difference between the MNGTs and LGGs. The most significant difference was seen with the maximum ADC value (ADC max) of the tumor where the values for LGGs were 1317 ± 314 whereas the values for MNGTs were 2134 ± 438. In both subsets of patients with MNGTs (DNET and Ganglioglioma), this difference was statistically significant (P = .015 and P = .0066, respectively). However, there was no significant difference between the ADC values of these subtypes of MNGTs.

Conclusion

The ADC values of MNGTs are significantly higher compared to LGGs and can be helpful in radiological demarcation of these two conditions. The high ADC of MNGTs may be attributable to the presence of large extracellular spaces and their cellularity, which is much lower than that of pure glial neoplasms.

Diffusion Weighted Imaging in Neuro-Oncology: Diagnosis, Post-Treatment Changes, and Advanced Sequences-An Updated Review

DWI is an imaging technique commonly used for the assessment of acute ischemia, inflammatory disorders, and CNS neoplasia. It has several benefits since it is a quick, easily replicable sequence that is widely used on many standard scanners. In addition to its normal clinical purpose, DWI offers crucial functional and physiological information regarding brain neoplasia and the surrounding milieu. A narrative review of the literature was conducted based on the PubMed database with the purpose of investigating the potential role of DWI in the neuro-oncology field. A total of 179 articles were included in the study.

Investigating the role of delayed contrast magnetic resonance imaging (MRI) to differentiate radiation necrosis from tumour recurrence in brain metastases after stereotactic radiosurgery

INTRODUCTION

The incidence of radionecrosis (RN) after stereotactic radiosurgery (SRS) to brain metastases is increasing. An overlap in the conventional MRI appearances of RN and tumour recurrence (TR) is diagnostically challenging. Delayed contrast MRI compares contrast enhancement over two time periods to create treatment response assessment maps (TRAMs). We aim to assess the utility of TRAMs in brain metastases patients.

METHODS

Delayed contrast MRI scans were performed on ten brain metastases patients, previously treated with SRS, who developed equivocal lesion(s) on routine MRI follow-up. T1-weighted images were obtained five minutes and 60-75 min after contrast injection, followed by Brain Lab software analysis to create TRAMs. TRAMs patterns were then compared with the patient's clinical status, subsequent imaging, and histology results.

RESULTS

We identified three regions on TRAMs: central, peripheral, and surrounding. Each region could be described either as contrast accumulation (red colour and representing non-tumour tissue) or contrast clearance (blue colour and representing tumour tissue). Our analysis demonstrated similarities in the TRAMs pattern between TR and RN, though to varying degrees.

CONCLUSION

In conclusion, the TRAMs appearances of RN and TR overlap. Our findings suggest that the previously-described correlation between contrast clearance and TR is at least partially attributable to more solid initial enhancement, rather than convincingly a difference in the underlying tissue properties, and the additional diagnostic value of TRAMs may be limited. Thus, further research on TRAMs is necessary prior to incorporating it into routine clinical management after SRS for brain metastases.

Molecular Imaging of Brain Tumors and Drug Delivery Using CEST MRI: Promises and Challenges

Chemical exchange saturation transfer (CEST) magnetic resonance imaging (MRI) detects molecules in their natural forms in a sensitive and non-invasive manner. This makes it a robust approach to assess brain tumors and related molecular alterations using endogenous molecules, such as proteins/peptides, and drugs approved for clinical use. In this review, we will discuss the promises of CEST MRI in the identification of tumors, tumor grading, detecting molecular alterations related to isocitrate dehydrogenase (IDH) and O-6-methylguanine-DNA methyltransferase (MGMT), assessment of treatment effects, and using multiple contrasts of CEST to develop theranostic approaches for cancer treatments. Promising applications include (i) using the CEST contrast of amide protons of proteins/peptides to detect brain tumors, such as glioblastoma multiforme (GBM) and low-grade gliomas; (ii) using multiple CEST contrasts for tumor stratification, and (iii) evaluation of the efficacy of drug delivery without the need of metallic or radioactive labels. These promising applications have raised enthusiasm, however, the use of CEST MRI is not trivial. CEST contrast depends on the pulse sequences, saturation parameters, methods used to analyze the CEST spectrum (i.e., Z-spectrum), and, importantly, how to interpret changes in CEST contrast and related molecular alterations in the brain. Emerging pulse sequence designs and data analysis approaches, including those assisted with deep learning, have enhanced the capability of CEST MRI in detecting molecules in brain tumors. CEST has become a specific marker for tumor grading and has the potential for prognosis and theranostics in brain tumors. With increasing understanding of the technical aspects and associated molecular alterations detected by CEST MRI, this young field is expected to have wide clinical applications in the near future.

Foundations of Multiparametric Brain Tumour Imaging Characterisation Using Machine Learning

The heterogeneity of brain tumours at the molecular, metabolic and structural levels poses significant challenge for accurate tissue characterisation. Artificial intelligence and radiomics have emerged as valuable tools to analyse quantitative features extracted from medical images which capture the complex microenvironment of brain tumours. In particular, a number of computational tools including machine learning algorithms have been proposed for image preprocessing, tumour segmentation, feature extraction, classification, and prognostic stratifications as well. In this chapter, we explore the fundamentals of multiparametric brain tumour characterisation, as an understanding of the strengths, limitations and applications of these tools allows clinicians to better develop and evaluate models with improved diagnostic and prognostic value in brain tumour patients.

Advanced Magnetic Resonance Imaging in Pediatric Glioblastomas

The shortly upcoming 5th edition of the World Health Organization Classification of Tumors of the Central Nervous System is bringing extensive changes in the terminology of diffuse high-grade gliomas (DHGGs). Previously "glioblastoma," as a descriptive entity, could have been applied to classify some tumors from the family of pediatric or adult DHGGs. However, now the term "glioblastoma" has been divested and is no longer applied to tumors in the family of pediatric types of DHGGs. As an entity, glioblastoma remains, however, in the family of adult types of diffuse gliomas under the insignia of "glioblastoma, IDH-wildtype." Of note, glioblastomas still can be detected in children when glioblastoma, IDH-wildtype is found in this population, despite being much more common in adults. Despite the separation from the family of pediatric types of DHGGs, what was previously labeled as "pediatric glioblastomas" still remains with novel labels and as new entities. As a result of advances in molecular biology, most of the previously called "pediatric glioblastomas" are now classified in one of the four family members of pediatric types of DHGGs. In this review, the term glioblastoma is still apocryphally employed mainly due to its historical relevance and the paucity of recent literature dealing with the recently described new entities. Therefore, "glioblastoma" is used here as an umbrella term in the attempt to encompass multiple entities such as astrocytoma, IDH-mutant (grade 4); glioblastoma, IDH-wildtype; diffuse hemispheric glioma, H3 G34-mutant; diffuse pediatric-type high-grade glioma, H3-wildtype and IDH-wildtype; and high grade infant-type hemispheric glioma. Glioblastomas are highly aggressive neoplasms. They may arise anywhere in the developing central nervous system, including the spinal cord. Signs and symptoms are non-specific, typically of short duration, and usually derived from increased intracranial pressure or seizure. Localized symptoms may also occur. The standard of care of "pediatric glioblastomas" is not well-established, typically composed of surgery with maximal safe tumor resection. Subsequent chemoradiation is recommended if the patient is older than 3 years. If younger than 3 years, surgery is followed by chemotherapy. In general, "pediatric glioblastomas" also have a poor prognosis despite surgery and adjuvant therapy. Magnetic resonance imaging (MRI) is the imaging modality of choice for the evaluation of glioblastomas. In addition to the typical conventional MRI features, i.e., highly heterogeneous invasive masses with indistinct borders, mass effect on surrounding structures, and a variable degree of enhancement, the lesions may show restricted diffusion in the solid components, hemorrhage, and increased perfusion, reflecting increased vascularity and angiogenesis. In addition, magnetic resonance spectroscopy has proven helpful in pre- and postsurgical evaluation. Lastly, we will refer to new MRI techniques, which have already been applied in evaluating adult glioblastomas, with promising results, yet not widely utilized in children.

DWI and ADC value versus ADC ratio in the characterization of solid renal masses: radiologic-pathologic correlation

Background

Renal masses are becoming an increasingly common finding on cross-sectional images. Characterization of the nature of the lesion either neoplastic or not, benign or malignant as well as further subtype characterization is becoming an important factor in determining management plan. The purpose of our study with to assess the sensitivity and specificity of both ADC mean value and ADC ratio in such characterization along with the calculation of different cutoff values to differentiate between different varieties, using pathological data as the main gold standard for diagnosis.

Results

Our study included 50 patients with a total of 72 masses. A final diagnosis was reached in 69 masses by pathological examination and three masses had clinical and laboratory signs of infection. We had a total of 49 malignant lesions (68%) and 23 benign lesions (32%). The ADC value of ccRCC (1.4 × 10−3 mm2/s) was significantly higher than all other renal masses. A cutoff ADC value of > 1.1 and a cutoff ADC ratio of > 0.56 can be used to differentiate between clear cell renal cell carcinoma and other lesions and an ADC value of < 0.8 and an ADC ratio of ≤ 0.56 to differentiate papillary renal cell carcinoma from other masses. There was no statistically significant ADC value to differentiate between benign and malignant lesions but a statistically significant ADC ratio (> 0.52) was reached.

Conclusion

ADC value and ADC ratio can be used as an adjunct tool in the characterization of different renal masses, with ADC ratio having a higher sensitivity, which can affect the prognosis and management of the patient.

Clinical applications of artificial intelligence and radiomics in neuro-oncology imaging

This article is a comprehensive review of the basic background, technique, and clinical applications of artificial intelligence (AI) and radiomics in the field of neuro-oncology. A variety of AI and radiomics utilized conventional and advanced techniques to differentiate brain tumors from non-neoplastic lesions such as inflammatory and demyelinating brain lesions. It is used in the diagnosis of gliomas and discrimination of gliomas from lymphomas and metastasis. Also, semiautomated and automated tumor segmentation has been developed for radiotherapy planning and follow-up. It has a role in the grading, prediction of treatment response, and prognosis of gliomas. Radiogenomics allowed the connection of the imaging phenotype of the tumor to its molecular environment. In addition, AI is applied for the assessment of extra-axial brain tumors and pediatric tumors with high performance in tumor detection, classification, and stratification of patient's prognoses.

A Fast Protocol for Multiparametric Characterisation of Diffusion in the Brain and Brain Tumours

Multi-parametric tissue characterisation is demonstrated using a 4-minute protocol based on diffusion trace acquisitions. Three diffusion regimes are covered simultaneously: pseudo-perfusion, Gaussian, and non-Gaussian diffusion. The clinical utility of this method for fast multi-parametric mapping for brain tumours is explored. A cohort of 17 brain tumour patients was measured on a 3T hybrid MR-PET scanner with a standard clinical MRI protocol, to which the proposed multi-parametric diffusion protocol was subsequently added. For comparison purposes, standard perfusion and a full diffusion kurtosis protocol were acquired. Simultaneous amino-acid (¹⁸F-FET) PET enabled the identification of active tumour tissue. The metrics derived from the proposed protocol included perfusion fraction, pseudo-diffusivity, apparent diffusivity, and apparent kurtosis. These metrics were compared to the corresponding metrics from the dedicated acquisitions: cerebral blood volume and flow, mean diffusivity and mean kurtosis. Simulations were carried out to assess the influence of fitting methods and noise levels on the estimation of the parameters. The diffusion and kurtosis metrics obtained from the proposed protocol show strong to very strong correlations with those derived from the conventional protocol. However, a bias towards lower values was observed. The pseudo-perfusion parameters showed very weak to weak correlations compared to their perfusion counterparts. In conclusion, we introduce a clinically applicable protocol for measuring multiple parameters and demonstrate its relevance to pathological tissue characterisation.

Conventional and advanced MRI evaluation of brain vascular malformations

Vascular malformations (VMs) of the central nervous system (CNS) include a wide range of pathological conditions related to intra and extracranial vessel abnormalities. Although some VMs show typical neuroimaging features, other VMs share and overlap pathological and neuroimaging features that hinder an accurate differentiation between them. Hence, it is not uncommon to misclassify different types of VMs under the general heading of arteriovenous malformations. Thorough knowledge of the imaging findings of each type of VM is mandatory to avoid these inaccuracies. Conventional MRI sequences, including MR angiography, have allowed the evaluation of CNS VMs without using ionizing radiation. Newer MRI techniques, such as susceptibility-weighted imaging, black blood sequences, arterial spin labeling, and 4D flow imaging, have an added value of providing physiopathological data in real time regarding the hemodynamics of VMs. Beyond MR images, new insights using 3D printed models are being incorporated as part of the armamentarium for a noninvasive evaluation of VMs. In this paper, we briefly review the pathophysiology of CNS VMs, focusing on the MRI findings that may be helpful to differentiate them. We discuss the role of each conventional and advanced MRI sequence for VMs assessment and provide some insights about the value of structured reports of 3D printing to evaluate VMs.

Diagnostic Accuracy of Machine Learning-Based Radiomics in Grading Gliomas: Systematic Review and Meta-Analysis

Purpose

This study aimed to estimate the diagnostic accuracy of machine learning- (ML-) based radiomics in differentiating high-grade gliomas (HGG) from low-grade gliomas (LGG) and to identify potential covariates that could affect the diagnostic accuracy of ML-based radiomic analysis in classifying gliomas.

Method

A primary literature search of the PubMed database was conducted to find all related literatures in English between January 1, 2009, and May 1, 2020, with combining synonyms for “machine learning,” “glioma,” and “radiomics.” Five retrospective designed original articles including LGG and HGG subjects were chosen. Pooled sensitivity, specificity, their 95% confidence interval, area under curve (AUC), and hierarchical summary receiver-operating characteristic (HSROC) models were obtained.

Result

The pooled sensitivity when diagnosing HGG was higher (96% (95% CI: 0.93, 0.98)) than the specificity when diagnosing LGG (90% (95% CI 0.85, 0.93)). Heterogeneity was observed in both sensitivity and specificity. Metaregression confirmed the heterogeneity in sample sizes (p=0.05), imaging sequence types (p=0.02), and data sources (p=0.01), but not for the inclusion of the testing set (p=0.19), feature extraction number (p=0.36), and selection of feature number (p=0.18). The results of subgroup analysis indicate that sample sizes of more than 100 and feature selection numbers less than the total sample size positively affected the diagnostic performance in differentiating HGG from LGG.

Conclusion

This study demonstrates the excellent diagnostic performance of ML-based radiomics in differentiating HGG from LGG.

Inflow-based vascular-space-occupancy (iVASO) might potentially predict IDH mutation status and tumor grade in diffuse cerebral gliomas

PURPOSE

The aim of the study is to assess the diagnostic performance of inflow-based vascular-space-occupancy (iVASO) MR imaging for differentiating glioblastomas (grade IV, GBM) and lower-grade diffuse gliomas (grade II and III, LGG) and its potential to predict IDH mutation status.

METHODS

One hundred and two patients with diffuse cerebral glioma (56 males; median age, 43.5 years) underwent iVASO and dynamic susceptibility contrast (DSC) MR imaging. The iVASO-derived arteriolar cerebral blood volume (CBVa), relative CBVa (rCBVa), and the DSC-derived relative cerebral blood volume (rCBV) were obtained, and these measurements were compared between the GBM group (n = 43) and the LGG group (n = 59) and between the IDH-mutation group (n = 54) and the IDH-wild group (n = 48).

RESULTS

Significant correlation was observed between rCBV and CBVa (P < 0.001) or rCBVa (P < 0.001). Both CBVa (P < 0.001) and rCBVa (P < 0.001) were higher in the GBM group. Both CBVa (P < 0.001) and rCBVa (P < 0.001) were lower in the IDH-mutation group compared to the IDH-wild group. Receiver operating characteristic analyses showed the area under curve (AUC) of 0.95 with CBVa and 0.97 with rCBVa in differentiating GBM from LGG. The AUCs were 0.82 and 0.85 for CBVa and rCBVa in predicting IDH gene status, respectively, which were lower than that of rCBV (AUC = 0.90). Combined rCBV and rCBVa significantly improved the diagnostic performance (AUC = 0.95).

CONCLUSIONS

iVASO MR imaging has the potential to predict IDH mutation and grade in glioma.

Feasibility of generalised diffusion kurtosis imaging approach for brain glioma grading

Purpose

An accurate differentiation of brain glioma grade constitutes an important clinical issue. Powerful non-invasive approach based on diffusion MRI has already demonstrated its feasibility in glioma grade stratification. However, the conventional diffusion tensor (DTI) and kurtosis imaging (DKI) demonstrated moderate sensitivity and performance in glioma grading. In the present work, we apply generalised DKI (gDKI) approach in order to assess its diagnostic accuracy and potential application in glioma grading.

Methods

Diffusion scalar metrics were obtained from 50 patients with different glioma grades confirmed by histological tests following biopsy or surgery. All patients were divided into two groups with low- and high-grade gliomas as grade II versus grades III and IV, respectively. For a comparison, trained radiologists segmented the brain tissue into three regions with solid tumour, oedema, and normal appearing white matter. For each region, we estimated the conventional and gDKI metrics including DTI maps.

Results

We found high correlations between DKI and gDKI metrics in high-grade glioma. Further, gDKI metrics enabled introduction of a complementary measure for glioma differentiation based on correlations between the conventional and generalised approaches. Both conventional and generalised DKI metrics showed quantitative maps of tumour heterogeneity and oedema behaviour. gDKI approach demonstrated largely similar sensitivity and specificity in low-high glioma differentiation as in the case of conventional DKI method.

Conclusion

The generalised diffusion kurtosis imaging enables differentiation of low- and high-grade gliomas at the same level as the conventional DKI. Additionally, gDKI exhibited higher sensitivity to tumour heterogeneity and tissue contrast between tumour and healthy tissue and, thus, may contribute as a complementary source of information on tumour differentiation.

Combined application of MRS and DWI can effectively predict cell proliferation and assess the grade of glioma: A prospective study

In order to assess combined application of MRS and DWI for prediction cell proliferation and grade diagnosis of glioma, We prospectively collected the Cho/Cr, Cho/NAA, Cr/NAA of MRS and tumor parenchyma ADC (ADC_T), contralateral mirror brain tissue ADC (ADC_H), rADC (rADC = ADC_T/ADC_H). According to postoperative pathology, the patients were divided into two groups: LGG group and HGG group, compared differences of age, gender, Ki67, MRS, DWI between two groups. Next, we analyzed the correlation between MRS, DWI and Ki67. On this basis, the sensitivity and specificity of MRS, DWI and MRS combined with DWI (MRS + DWI) in diagnosis of glioma grade were evaluated. The differences of Ki67, Cho/Cr, Cho/NAA, Cr/NAA, ADC_T, rADC between LGG group and HGG group were statistically significant (p = 0.000, 0.000, 0.000, 0.008, 0.000, and 0.000 respectively). From ROC curve, area under the curve (AUC), sensitivity and specificity of Cho/Cr, Cho/NAA, Cr/NAA, ADC_T, rADC, PRE (MRS + DWI) were (0.901, 86.7%, 85.7%), (0.876, 80.0%, 82.1%), (0.704, 63.3%, 71.4%), (0.862, 82.1%, 83.3%), (0.820, 75.0%, 76.7%), (0.920, 86.7%, 89.3%), respectively. Fisher's linear discriminant functions results suggest: Y₁ = -20.447 + 3.46•X₁ + 17.141•X₂ (LGG), Y₂ = -19.415 + 4.828•X₁ + 14.543•X₂ (HGG). Our study suggested that MRS and DWI can effectively predict cell proliferation preoperative. MRS combined with DWI can further improve sensitivity and specificity in assessing the grade of glioma.

Differentiation between high-grade gliomas and solitary brain metastases: a comparison of five diffusion-weighted MRI models

Background

To compare the diagnostic performance of neurite orientation dispersion and density imaging (NODDI), mean apparent propagator magnetic resonance imaging (MAP-MRI), diffusion kurtosis imaging (DKI), diffusion tensor imaging (DTI) and diffusion-weighted imaging (DWI) in distinguishing high-grade gliomas (HGGs) from solitary brain metastases (SBMs).

Methods

Patients with previously untreated, histopathologically confirmed HGGs (n = 20) or SBMs (n = 21) appearing as a solitary and contrast-enhancing lesion on structural MRI were prospectively recruited to undergo diffusion-weighted MRI. DWI data were obtained using a q-space Cartesian grid sampling procedure and were processed to generate parametric maps by fitting the NODDI, MAP-MRI, DKI, DTI and DWI models. The diffusion metrics of the contrast-enhancing tumor and peritumoral edema were measured. Differences in the diffusion metrics were compared between HGGs and SBMs, followed by receiver operating characteristic (ROC) analysis and the Hanley and McNeill test to determine their diagnostic performances.

Results

NODDI-based isotropic volume fraction (V_iso) and orientation dispersion index (ODI); MAP-MRI-based mean-squared displacement (MSD) and q-space inverse variance (QIV); DKI-generated radial, mean diffusivity and fractional anisotropy (RD_k, MD_k and FA_k); and DTI-generated radial, mean diffusivity and fractional anisotropy (RD, MD and FA) of the contrast-enhancing tumor were significantly different between HGGs and SBMs (p < 0.05). The best single discriminative parameters of each model were V_iso, MSD, RD_k and RD for NODDI, MAP-MRI, DKI and DTI, respectively. The AUC of V_iso (0.871) was significantly higher than that of MSD (0.736), RD_k (0.760) and RD (0.733) (p < 0.05).

Conclusion

NODDI outperforms MAP-MRI, DKI, DTI and DWI in differentiating between HGGs and SBMs. NODDI-based V_iso has the highest performance.

Precise discrimination between meningiomas and schwannomas using time-to-signal intensity curves and percentage signal recoveries obtained from dynamic susceptibility perfusion imaging

BACKGROUND AND PURPOSE

Meningiomas and schwannomas are common extra-axial brain tumors. Discrimination is challenging in some locations when characteristic imaging features are absent. This study investigated the accuracy of percentage signal recoveries obtained from dynamic susceptibility contrast perfusion imaging (DSC-PI) in discriminating meningiomas and schwannomas.

MATERIAL AND METHODS

Retrospective database research was conducted. Sixty nine meningioma and 15 schwannoma having DSC-PI between January 2016 and February 2020 were included. Time to signal intensity curves (TSIC) were analyzed and grouped as T1-dominant leakage, T2*-dominant leakage and return to baseline. Relative cerebral blood volume (rCBV), relative mean transit time (rMTT), percentage signal recovery 1 (PSR 1) and PSR 2 values were calculated. The differences between the groups were investigated. Receiver operating characteristic curves were operated.

RESULTS

rCBV, rMTT, PSR 1 and PSR 2 values were statistically different between meningiomas and schwannomas. PSR 2 provided the best discrimination. With the cut off value of 1.08 for PSR 2, meningiomas and schwannomas were differentiated with 95.7% sensitivity and 93.3% specificity. TSICs were also different between two groups. Most of meningiomas showed T2*-dominant leakage (78.2%), whereas most of shwannomas showed T1-dominant leakage (93.3%).

CONCLUSION

DSC-PI is a useful imaging tool for non-invasive discrimination of meningiomas and schwannomas. Particularly, percentage signal recoveries discriminates meningiomas and schwannomas with high sensitivity and specificity.

Machine Learning in Meningioma MRI: Past to Present. A Narrative Review

Meningioma is one of the most frequent primary central nervous system tumors. While magnetic resonance imaging (MRI), is the standard radiologic technique for provisional diagnosis and surveillance of meningioma, it nevertheless lacks the prima facie capacity in determining meningioma biological aggressiveness, growth, and recurrence potential. An increasing body of evidence highlights the potential of machine learning and radiomics in improving the consistency and productivity and in providing novel diagnostic, treatment, and prognostic modalities in neuroncology imaging. The aim of the present article is to review the evolution and progress of approaches utilizing machine learning in meningioma MRI-based sementation, diagnosis, grading, and prognosis. We provide a historical perspective on original research on meningioma spanning over two decades and highlight recent studies indicating the feasibility of pertinent approaches, including deep learning in addressing several clinically challenging aspects. We indicate the limitations of previous research designs and resources and propose future directions by highlighting areas of research that remain largely unexplored. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY STAGE: 2.

Deep Learning-based Noise Reduction for Fast Volume Diffusion Tensor Imaging: Assessing the Noise Reduction Effect and Reliability of Diffusion Metrics

To assess the feasibility of a denoising approach with deep learning-based reconstruction (dDLR) for fast volume simultaneous multi-slice diffusion tensor imaging of the brain, noise reduction effects and the reliability of diffusion metrics were evaluated with 20 patients. Image noise was significantly decreased with dDLR. Although fractional anisotropy (FA) of deep gray matter was overestimated when the number of image acquisitions was one (NAQ1), FA in NAQ1 with dDLR became closer to that in NAQ5.

Histogram Analysis of ADC Maps for Differentiating Brain Metastases From Different Histological Types of Lung Cancers

PURPOSE

Our study aimed to investigate the role of histogram analysis derived from apparent diffusion coefficient (ADC) maps in brain metastases (BMs) from lung cancer for differentiating histological subtype.

METHODS

A total of 61 BMs (45 non-small cell lung cancer [NSCLC] comprising 32 adenocarcinoma [AC], 13 squamous cell carcinoma [SCC], and 16 small-cell lung cancer [SCLC]) in 50 patients with histopathologically confirmed lung cancer were retrospectively included in this study. Pretreatment cranial diffusion-weighted imaging was performed, and the corresponding ADC maps were generated. Regions of interest were drawn on solid components of the BM on all slices of the ADC maps to obtain parameters, including ADC_max, ADC_mean, ADC_min, ADC_median, ADC_range, skewness, kurtosis, entropy, ADC₁₀, ADC₂₅, ADC₇₅, and ADC₉₀. Apparent diffusion coefficient histogram parameters were compared among histological type groups. Kruskal-Wallis, Mann-Whitney U, chi-square tests, and receiver-operating characteristic (ROC) curve were used for statistical assessment.

RESULTS

ADC_min, ADC₁₀, and ADC₂₅ were found to be significantly different among AC, SCC, and SCLC groups; these parameters were higher for AC group, moderate for SCC group, and significantly lower for SCLC group. Skewness and kurtosis were not significantly different among all groups. The ROC analysis for differentiating BMs of NSCLC from SCLC showed that ADC₂₅ achieved the highest area under the curve at 0.922 with 93.02% sensitivity and 81.25% specificity.

CONCLUSION

Apparent diffusion coefficient histogram analysis of BMs from lung cancer has significant prognostic value in differentiating histological subtypes of lung cancer.

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.

Diffusion-Weighted Imaging in Oncology: An Update

To date, diffusion weighted imaging (DWI) is included in routine magnetic resonance imaging (MRI) protocols for several cancers. The real additive role of DWI lies in the "functional" information obtained by probing the free diffusivity of water molecules into intra and inter-cellular spaces that in tumors mainly depend on cellularity. Although DWI has not gained much space in some oncologic scenarios, this non-invasive tool is routinely used in clinical practice and still remains a hot research topic: it has been tested in almost all cancers to differentiate malignant from benign lesions, to distinguish different malignant histotypes or tumor grades, to predict and/or assess treatment responses, and to identify residual or recurrent tumors in follow-up examinations. In this review, we provide an up-to-date overview on the application of DWI in oncology.

Leveraging multi-shell diffusion for studies of brain development in youth and young adulthood

•

Multi-shell imaging sequences may improve sensitivity to developmental effects.

•

Models that leverage multi-shell information are often less sensitive to the confounding effects of motion.

•

Multi-shell sequences and models that leverage this data may be of particular utility for studying the developing brain.

Diffusion weighted imaging (DWI) has advanced our understanding of brain microstructure evolution over development. Recently, the use of multi-shell diffusion imaging sequences has coincided with advances in modeling the diffusion signal, such as Neurite Orientation Dispersion and Density Imaging (NODDI) and Laplacian-regularized Mean Apparent Propagator MRI (MAPL). However, the relative utility of recently-developed diffusion models for understanding brain maturation remains sparsely investigated. Additionally, despite evidence that motion artifact is a major confound for studies of development, the vulnerability of metrics derived from contemporary models to in-scanner motion has not been described. Accordingly, in a sample of 120 youth and young adults (ages 12–30) we evaluated metrics derived from diffusion tensor imaging (DTI), NODDI, and MAPL for associations with age and in-scanner head motion at multiple scales. Specifically, we examined mean white matter values, white matter tracts, white matter voxels, and connections in structural brain networks. Our results revealed that multi-shell diffusion imaging data can be leveraged to robustly characterize neurodevelopment, and demonstrate stronger age effects than equivalent single-shell data. Additionally, MAPL-derived metrics were less sensitive to the confounding effects of head motion. Our findings suggest that multi-shell imaging data and contemporary modeling techniques confer important advantages for studies of neurodevelopment.

Detection of Glioblastoma Subclinical Recurrence Using Serial Diffusion Tensor Imaging

Glioblastoma is an aggressive brain tumor with a propensity for intracranial recurrence. We hypothesized that tumors can be visualized with diffusion tensor imaging (DTI) before they are detected on anatomical magnetic resonance (MR) images. We retrospectively analyzed serial MR images from 30 patients, including the DTI and T1-weighted images at recurrence, at 2 months and 4 months before recurrence, and at 1 month after radiation therapy. The diffusion maps and T1 images were deformably registered longitudinally. The recurrent tumor was manually segmented on the T1-weighted image and then applied to the diffusion maps at each time point to collect mean FA, diffusivities, and neurite density index (NDI) values, respectively. Group analysis of variance showed significant changes in FA (p = 0.01) and NDI (p = 0.0015) over time. Pairwise t tests also revealed that FA and NDI at 2 months before recurrence were 11.2% and 6.4% lower than those at 1 month after radiation therapy (p < 0.05), respectively. Changes in FA and NDI were observed 2 months before recurrence, suggesting that progressive microstructural changes and neurite density loss may be detectable before tumor detection in anatomical MR images. FA and NDI may serve as non-contrast MR-based biomarkers for detecting subclinical tumors.

Histogram Analysis of Diffusion Weighted Imaging in Low-Grade Gliomas: in vivo Characterization of Tumor Architecture and Corresponding Neuropathology

Background: Low-grade gliomas (LGG) in adults are usually slow growing and frequently asymptomatic brain tumors, originating from glial cells of the central nervous system (CNS). Although regarded formally as “benign” neoplasms, they harbor the potential of malignant transformation associated with high morbidity and mortality. Their complex and unpredictable tumor biology requires a reliable and conclusive presurgical magnetic resonance imaging (MRI). A promising and emerging MRI approach in this context is histogram based apparent diffusion coefficient (ADC) profiling, which recently proofed to be capable of providing prognostic relevant information in different tumor entities. Therefore, our study investigated whether histogram profiling of ADC distinguishes grade I from grade II glioma, reflects the proliferation index Ki-67, as well as the IDH (isocitrate dehydrogenase) mutation and MGMT (methylguanine-DNA methyl-transferase) promotor methylation status.

Material and Methods: Pre-treatment ADC volumes of 26 LGG patients were used for histogram-profiling. WHO-grade, Ki-67 expression, IDH mutation, and MGMT promotor methylation status were evaluated. Comparative and correlative statistics investigating the association between histogram-profiling and neuropathology were performed.

Results: Almost the entire ADC profile (p25, p75, p90, mean, median) was significantly lower in grade II vs. grade I gliomas. Entropy, as second order histogram parameter of ADC volumes, was significantly higher in grade II gliomas compared with grade I gliomas. Mean, maximum value (ADCmax) and the percentiles p10, p75, and p90 of ADC histogram were significantly correlated with Ki-67 expression. Furthermore, minimum ADC value (ADCmin) was significantly associated with MGMT promotor methylation status as well as ADC entropy with IDH-1 mutation status.

Conclusions: ADC histogram-profiling is a valuable radiomic approach, which helps differentiating tumor grade, estimating growth kinetics and probably prognostic relevant genetic as well as epigenetic alterations in LGG.

Atypical and malignant canine intracranial meningiomas may have lower apparent diffusion coefficient values than benign tumors

Canine intracranial meningiomas can be graded based on histological classification as benign (grade I), atypical (grade II), and anaplastic or malignant (grade III). In people, grade II/III meningiomas behave more aggressively, have a higher potential for recurrence after surgical resection, and have lower apparent diffusion coefficient (ADC) values with diffusion weighted imaging (DWI). In this retrospective analytical cross-sectional study, 42 dogs had ADC values quantified in an attempt to differentiate tumor histologic grade. Our hypothesis was that ADC values would be significantly lower in grade II and III versus grade I meningiomas in dogs. On each ADC image, a polygonal region of interest (ROI) was hand-drawn along the lesion's periphery, excluding fluid-filled and hemorrhagic regions. Mean ADC value (ADC_mean ) and minimum ADC value (ADC_min ) were calculated. Additionally, two smaller, ovoid ROI were drawn within the lesion with mean ADC calculated (ADC_{mean sR} and ADC_{min sR} ). Normalized ADC values using white matter were also calculated (ADC_n and ADC_{n sR} ). Grades of each tumor were assigned based on histopathology review. Association between ADC parameters and histological grade was tested by means of two-sample t-tests. There were 14 grade I (33.3%), 25 grade II (59.5%), and three grade III (7.2%) meningiomas. ADC_{mean sR} and ADC_{min sR} were significantly lower when comparing grade II/III to grade I (P < .05). Grade II tumors had significantly lower ADC_mean , ADC_{mean sR} , ADC_{min sR} , ADC_n , and ADC_{n sR} than grade I meningiomas. This preliminary study supports the potential of ADC values to help predict the histological grade of intracranial meningiomas in dogs.

Combination value of diffusion‑weighted imaging and dynamic susceptibility contrast‑enhanced MRI in astrocytoma grading and correlation with GFAP, Topoisomerase IIα and MGMT

The present study aimed to investigate the value of diffusion-weighted imaging (DWI) combined with dynamic susceptibility contrast-enhanced (DSC) magnetic resonance imaging (MRI) scans in astrocytoma grading, and correlated MRI scan parameters of values of apparent diffusion coefficient (ADC) and relative cereberal blood volume (rCBV) with the immunohistochemical (IHC) indices of glial fibrillary acidic protein (GFAP), topoisomerase IIα (Topo IIα) and O 6-methylguanine-DNA methyltransferase (MGMT). A total of 123 patients with pathologically confirmed astrocytomas of differing grades underwent DWI and DSC scans. The values of the ADC and relative cerebral blood volume rCBV were compared with the grade II–IV astrocytomas. Receiver operating characteristic curves were used to compare astrocytoma grading efficiency of ADC, rCBV and the combination of the two values. The parameters of ADC and rCBV with GFAP, Topo IIα and MGMT indices were then correlated. The differences in ADC values were significant between the grades II, III and IV astrocytomas, and the rCBV values for grades II, III and IV were also significant. The combination of DWI and DSC demonstrated the highest values for area under curve in identifying grades II and III, and identifying grades III and IV, respectively. GFAP displayed a positive correlation with ADC and a negative correlation with rCBV. Topo IIα exhibited a negative correlation with ADC, and a positive correlation with rCBV. No correlation was observed between MGMT and ADC or rCBV. The combined application of DWI and DSC may increase astrocytoma grading accuracy. Values of ADC and rCBV exhibit certain correlations with IHC indices, and may predict degree of malignancy of astrocytoma.

Accuracy of apparent diffusion coefficients and enhancement ratios on magnetic resonance imaging in differentiating primary cerebral lymphomas from glioblastoma

BACKGROUND AND PURPOSE

This study aimed to determine the accuracy of apparent diffusion coefficient (ADC) and enhancement ratio (ER) in discriminating primary cerebral lymphomas (PCL) and glioblastomas.

MATERIALS AND METHODS

Circular regions of interest were randomly placed centrally within the largest solid-enhancing area of all lymphomas and glioblastomas on both post-contrast T1-weighted images and corresponding ADC maps. Regions of interest were also drawn in the contralateral hemisphere to obtain enhancement and ADC values of normal-appearing white matter. This helped us to calculate the ER and ADC ratio.

RESULTS

Mean enhancement and ADC (mm²/s) values for PCL were 2220.56 ± 2948.30 and 712.00 ± 137.87, respectively. Mean enhancement and ADC values for glioblastoma were 1537.07 ± 1668.33 and 1037.93 ± 280.52, respectively. Differences in ADC values, ratios and ERs were all statistically significant between the two groups (p < 0.05). ADC values correctly predicted 71.4% of the lesions as glioblastoma and 83.3% as PCL (area under the curve (AUC) = 0.86 on receiver operating characteristic curve analysis). ADC ratios correctly predicted 85.7% of the lesions as glioblastoma and 100% as PCL (AUC = 0.93). ERs correctly predicted 71.4% of the lesions as glioblastoma and 88.9% as PCL (AUC = 0.92). The combination of ADC ratio and ER correctly predicted 100% tumour type in both patient subgroups.

CONCLUSIONS

ADC values, ADC ratios and ERs may serve as useful variables to distinguish PCL from glioblastoma. The combination of ADC ratio and ER yielded the best results in identification of both patient subgroups.

Application of MR morphologic, diffusion tensor, and perfusion imaging in the classification of brain tumors using machine learning scheme

PURPOSE

While MRI is the modality of choice for the assessment of patients with brain tumors, differentiation between various tumors based on their imaging characteristics might be challenging due to overlapping imaging features. The purpose of this study was to apply a machine learning scheme using basic and advanced MR sequences for distinguishing different types of brain tumors.

METHODS

The study cohort included 141 patients (41 glioblastoma, 38 metastasis, 50 meningioma, and 12 primary central nervous system lymphoma). A computer-assisted classification scheme, combining morphologic MRI, perfusion MRI, and DTI metrics, was developed and used for tumor classification. The proposed multistep scheme consists of pre-processing, ROI definition, features extraction, feature selection, and classification. Feature subset selection was performed using support vector machines (SVMs). Classification performance was assessed by leave-one-out cross-validation. Given an ROI, the entire classification process was done automatically via computer and without any human intervention.

RESULTS

A binary hierarchical classification tree was chosen. In the first step, selected features were chosen for distinguishing glioblastoma from the remaining three classes, followed by separation of meningioma from metastasis and PCNSL, and then to discriminate PCNSL from metastasis. The binary SVM classification accuracy, sensitivity and specificity for glioblastoma, metastasis, meningiomas, and primary central nervous system lymphoma were 95.7, 81.6, and 91.2%; 92.7, 95.1, and 93.6%; 97, 90.8, and 58.3%; and 91.5, 90, and 96.9%, respectively.

CONCLUSION

A machine learning scheme using data from anatomical and advanced MRI sequences resulted in high-performance automatic tumor classification algorithm. Such a scheme can be integrated into clinical decision support systems to optimize tumor classification.

Probabilistic independent component analysis of dynamic susceptibility contrast perfusion MRI in metastatic brain tumors

PURPOSE

To identify clinically relevant magnetic resonance imaging (MRI) features of different types of metastatic brain lesions, including standard anatomical, diffusion weighted imaging (DWI) and dynamic susceptibility contrast (DSC) perfusion MRI.

METHODS

MRI imaging was retrospectively assessed on one hundred and fourteen (N = 114) brain metastases including breast (n = 27), non-small cell lung cancer (NSCLC, n = 43) and 'other' primary tumors (n = 44). Based on 114 patient's MRI scans, a total of 346 individual contrast enhancing tumors were manually segmented. In addition to tumor volume, apparent diffusion coefficients (ADC) and relative cerebral blood volume (rCBV) measurements, an independent component analysis (ICA) was performed with raw DSC data in order to assess arterio-venous components and the volume of overlap (AVOL) relative to tumor volume, as well as time to peak (TTP) of T₂* signal from each component.

RESULTS

Results suggests non-breast or non-NSCLC ('other') tumors had higher volume compare to breast and NSCLC patients (p = 0.0056 and p = 0.0003, respectively). No differences in median ADC or rCBV were observed across tumor types; however, breast and NSCLC tumors had a significantly higher "arterial" proportion of the tumor volume as indicated by ICA (p = 0.0062 and p = 0.0018, respectively), while a higher "venous" proportion were prominent in breast tumors compared with NSCLC (p = 0.0027) and 'other' lesions (p = 0.0011). The AVOL component was positively related to rCBV in all groups, but no correlation was found for arterial and venous components with respect to rCBV values. Median time to peak of arterial and venous components were 8.4 s and 12.6 s, respectively (p < 0.0001). No difference was found in arterial or venous TTP across groups.

CONCLUSIONS

Advanced ICA-derived component analysis demonstrates perfusion differences between metastatic brain tumor types that were not observable with classical ADC and rCBV measurements. These results highlight the complex relationship between brain tumor vasculature characteristics and the site of primary tumor diagnosis.

An evidence-based approach to assess the accuracy of intravoxel incoherent motion imaging for the grading of brain tumors

BACKGROUND

Differentiation of high-grade gliomas (HGGs) and low-grade gliomas (LGGs) is an important clinical problem because treatment strategies vary greatly. This study was performed to investigate the potential diagnostic value of incoherent intravoxel motion imaging (IVIM) to distinguish HGG from LGG by meta-analysis.

METHODS

A computerized search of the literature was performed using the free-access PubMed database, Web of Science, and Chinese biomedical database, and relevant articles until September 18, 2018 that used IVIM to distinguish HGG from LGG were included. All analyses were performed using Review Manager 5.3 and Stata. Mean difference (MD) at 95% confidence interval (CI) of the apparent diffusion coefficient (ADC), diffusion coefficient value (D), perfusion fraction value (f), and perfusion coefficient value (D*) were summarized.

RESULTS

Nine studies were used for general data pooling. In the tumor parenchyma (TP) regions, subgroup analysis revealed D* in HGG is higher than in LGG (MD = 1.19, P = .002), and D in HGG is lower than in LGG (MD = -1.06, P = .001). However, no significant difference in f (MD = 0.89, P = .056) was detected between HGG and LGG. In the white matter regions, HGG had higher D* (MD = 0.76, P = .04) compared with LGG, while no marked differences between the D value (P = .07) and f (P = .09) values.

CONCLUSION

The present meta-analysis shows that the ADC, D, and D* values derived from IVIM may be useful in differentiating HGG from LGG. Considering the small sample of this study, we need to be cautious when interpreting the results of this study. Other prospective and large-sample randomized controlled trials were needed to establish the value of IVIM in differentiating HGG from LGG.

Pyogenic brain abscess, ventriculitis and diffuse meningitis with fatal outcome in an adult: Radiologic-pathologic correlation☆,

Rupture of brain abscesses with evolution into ventriculitis with meningitis may result in sudden and dramatic worsening of the clinical situation. We present a 57-year-old man with such an event and fatal outcome. Multiple imaging modalities including computed tomography and advanced magnetic resonance imaging were correlated with gross specimen and histologic images. The differential diagnosis of multiple lesions with ring enhancement and prominent perifocal edema includes mainly infectious and neoplastic processes, such as brain abscess, metastasis, and multicentric glioblastoma. Pyogenic ventriculitis is an uncommon manifestation of severe intracranial infection that might be clinically obscure. We discuss the characteristic magnetic resonance findings of brain abscess and its complications, including meningitis and ventriculitis with emphasis on the role of diffusion-weighted and fluid-attenuated inversion recovery imaging.

Primary Skull Base Lymphoma: Manifestations and Clinical Outcomes of a Great Imitator

Objectives Primary skull base lymphoma (PSBL) represents a rare manifestation of extranodal lymphoma. Presenting with nonspecific symptomatology and imaging findings, PSBL often masquerades as more common cranial base pathology and thus can present a diagnostic challenge. The objectives of this study were to characterize the manifestations and clinical outcomes of PSBL. Study Design Case series and chart review. Setting Tertiary referral center. Subjects and Methods Review of 48 patients with PSBL treated between 1994 and 2016. Results The median age at diagnosis was 60 years. Diffuse large B-cell lymphoma comprised the most common subtype (56%). Presenting symptoms included diplopia (52%), trigeminal hypesthesia (38%), headache (29%), facial nerve weakness (25%), B-symptoms (fevers, night sweats, and/or weight loss) (25%), and hearing loss (21%). Discrete lesions commonly mimicked meningioma, schwannoma, and nasopharyngeal carcinoma. Diffuse lesions imitated neurosarcoidosis, granulomatosis with polyangiitis, and carcinomatosis meningitis. Only 26% of all initial clinical evaluations suspected lymphoma. The combination of restricted diffusion on diffusion-weighted imaging, an absence of flow voids or surrounding hyperostosis, bony erosion and/or marrow signal replacement, and heterogeneous contrast enhancement facilitated delineation between PSBL and common skull base pathology. Cerebrospinal fluid (CSF) analysis was diagnostic of lymphoma in 24% of cases. Lymphoma within CSF portended significantly worse overall survival (85% vs 18% at 3-years; P < .001). Conclusion A history significant for multiple cranial nerve palsies, B-symptoms, and imaging findings inconsistent with common skull base pathology should raise suspicion for PSBL. CSF analysis in the setting of PSBL can provide diagnostic and prognostic value for patients.

Radiomics strategy for glioma grading using texture features from multiparametric MRI

BACKGROUND

Accurate glioma grading plays an important role in the clinical management of patients and is also the basis of molecular stratification nowadays.

PURPOSE/HYPOTHESIS

To verify the superiority of radiomics features extracted from multiparametric MRI to glioma grading and evaluate the grading potential of different MRI sequences or parametric maps.

STUDY TYPE

Retrospective; radiomics.

POPULATION

A total of 153 patients including 42, 33, and 78 patients with Grades II, III, and IV gliomas, respectively.

FIELD STRENGTH/SEQUENCE

3.0T MRI/T₁ -weighted images before and after contrast-enhanced, T₂ -weighted, multi-b-value diffusion-weighted and 3D arterial spin labeling images.

ASSESSMENT

After multiparametric MRI preprocessing, high-throughput features were derived from patients' volumes of interests (VOIs). The support vector machine-based recursive feature elimination was adopted to find the optimal features for low-grade glioma (LGG) vs. high-grade glioma (HGG), and Grade III vs. IV glioma classification tasks. Then support vector machine (SVM) classifiers were established using the optimal features. The accuracy and area under the curve (AUC) was used to assess the grading efficiency.

STATISTICAL TESTS

Student's t-test or a chi-square test were applied on different clinical characteristics to confirm whether intergroup significant differences exist.

RESULTS

Patients' ages between LGG and HGG groups were significantly different (P < 0.01). For each patient, 420 texture and 90 histogram parameters were derived from 10 VOIs of multiparametric MRI. SVM models were established using 30 and 28 optimal features for classifying LGGs from HGGs and grades III from IV, respectively. The accuracies/AUCs were 96.8%/0.987 for classifying LGGs from HGGs, and 98.1%/0.992 for classifying grades III from IV, which were more promising than using histogram parameters or using the single sequence MRI.

DATA CONCLUSION

Texture features were more effective for noninvasively grading gliomas than histogram parameters. The combined application of multiparametric MRI provided a higher grading efficiency. The proposed radiomic strategy could facilitate clinical decision-making for patients with varied glioma grades.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;48:1518-1528.

Brain Tumor Characterization Using Multibiometric Evaluation of MRI

The aim was to evaluate volume, diffusion, and perfusion metrics for better presurgical differentiation between high-grade gliomas (HGG), low-grade gliomas (LGG), and metastases (MET). For this retrospective study, 43 patients with histologically verified intracranial HGG (n = 18), LGG (n = 10), and MET (n = 15) were chosen. Preoperative magnetic resonance data included pre- and post-gadolinium contrast-enhanced T1-weighted fluid-attenuated inversion recover, cerebral blood flow (CBF), cerebral blood volume (CBV), fractional anisotropy, and apparent diffusion coefficient maps used for quantification of magnetic resonance biometrics by manual delineation of regions of interest. A binary logistic regression model was applied for multiparametric analysis and receiver operating characteristic (ROC) analysis. Statistically significant differences were found for normalized-ADC-tumor (nADC-T), normalized-CBF-tumor (nCBF-T), normalized-CBV-tumor (nCBV-T), and normalized-CBF-edema (nCBF-E) between LGG and HGG, and when these metrics were combined, HGG could be distinguished from LGG with a sensitivity and specificity of 100%. The only metric to distinguish HGG from MET was the normalized-ADC-E with a sensitivity of 68.8% and a specificity of 80%. LGG can be distinguished from MET by combining edema volume (Vol-E), Vol-E/tumor volume (Vol-T), nADC-T, nCBF-T, nCBV-T, and nADC-E with a sensitivity of 93.3% and a specificity of 100%. The present study confirms the usability of a multibiometric approach including volume, perfusion, and diffusion metrics in differentially diagnosing brain tumors in preoperative patients and adds to the growing body of evidence in the clinical field in need of validation and standardization.

Principles of diffusion kurtosis imaging and its role in early diagnosis of neurodegenerative disorders

Pathology of neurodegenerative diseases can be correlated with intra-neuronal as well as extracellular changes which lead to neuronal degeneration. The central nervous system (CNS) is a complex structure comprising of many biological barriers. These microstructural barriers might be affected by a variety of pathological processes. Specifically, changes in the brain tissue's microstructure affect the diffusion of water which can be assessed non-invasively by diffusion weighted (DW) magnetic resonance imaging (MRI) techniques. Diffusion tensor imaging (DTI) is a diffusion MRI technique that considers diffusivity as a Gaussian process, i.e. does not account for any diffusion hindrance. However, environment of the brain tissues is characterized by a non-Gaussian diffusion. Therefore, diffusion kurtosis imaging (DKI) was developed as an extension of DTI method in order to quantify the non-Gaussian distribution of water diffusion. This technique represents a promising approach for early diagnosis of neurodegenerative diseases when the neurodegenerative process starts. Hence, the purpose of this article is to summarize the ongoing clinical and preclinical research on Parkinson's, Alzheimer's and Huntington diseases, using DKI and to discuss the role of this technique as an early stage biomarker of neurodegenerative conditions.

An improved model for prostate diffusion incorporating the results of Monte Carlo simulations of diffusion in the cellular compartment

The purpose of this work was to refine a previously published model of prostate diffusion by incorporating improved estimates of cellular diffusivity obtained by Monte Carlo simulation. Stromal and epithelial cell size and intracellular volume fraction in different grades of cancer were determined from histological images. Diffusion in different mixtures of cells, corresponding to different tumor grades, was simulated and cellular apparent diffusion coefficient and kurtosis values determined. These values were incorporated into the previously published model of prostate diffusion and model predictions compared with values found in the literature. Stromal cell radius and intracellular volume fraction were 3.74 ± 0.96 μm and 13 ± 3% respectively in normal peripheral zone (PZ), and were similar in all grades of cancer. Epithelial cell radius and intracellular volume fraction were 3.40 ± 0.15 μm and 45 ± 5% respectively in normal PZ, rising to 4.75 ± 0.20 μm and 70 ± 8% in high grade cancer. Cellular apparent diffusion coefficient and kurtosis were 1.02 μm² ms^-1 and 0.58 respectively in normal PZ, and 0.61 μm² ms^-1 and 1.15 in high grade cancer (variation in simulation values are less than 0.1%). Agreement between model predictions and measurements were good, with a mean square error of 0.22 μm² ms^-1 . Incorporation of cellular diffusion coefficient and kurtosis values obtained by Monte Carlo simulation into a model of prostate diffusion gives good agreement with published results.

Comparison of 18F-FET PET and perfusion-weighted MRI for glioma grading: a hybrid PET/MR study

PURPOSE

Both perfusion-weighted MR imaging (PWI) and O-(2-¹⁸F-fluoroethyl)-L-tyrosine PET (¹⁸F-FET) provide grading information in cerebral gliomas. The aim of this study was to compare the diagnostic value of ¹⁸F-FET PET and PWI for tumor grading in a series of patients with newly diagnosed, untreated gliomas using an integrated PET/MR scanner.

METHODS

Seventy-two patients with untreated gliomas [22 low-grade gliomas (LGG), and 50 high-grade gliomas (HGG)] were investigated with ¹⁸F-FET PET and PWI using a hybrid PET/MR scanner. After visual inspection of PET and PWI maps (rCBV, rCBF, MTT), volumes of interest (VOIs) with a diameter of 16 mm were centered upon the maximum of abnormality in the tumor area in each modality and the contralateral unaffected hemisphere. Mean and maximum tumor-to-brain ratios (TBR_mean, TBR_max) were calculated. In addition, Time-to-Peak (TTP) and slopes of time-activity curves were calculated for ¹⁸F-FET PET. Diagnostic accuracies of ¹⁸F-FET PET and PWI for differentiating low-grade glioma (LGG) from high-grade glioma (HGG) were evaluated by receiver operating characteristic analyses (area under the curve; AUC).

RESULTS

The diagnostic accuracy of ¹⁸F-FET PET and PWI to discriminate LGG from HGG was similar with highest AUC values for TBR_mean and TBR_max of ¹⁸F-FET PET uptake (0.80, 0.83) and for TBR_mean and TBR_max of rCBV (0.80, 0.81). In case of increased signal in the tumor area with both methods (n = 32), local hot-spots were incongruent in 25 patients (78%) with a mean distance of 10.6 ± 9.5 mm. Dynamic FET PET and combination of different parameters did not further improve diagnostic accuracy.

CONCLUSIONS

Both ¹⁸F-FET PET and PWI discriminate LGG from HGG with similar diagnostic performance. Regional abnormalities in the tumor area are usually not congruent indicating that tumor grading by ¹⁸F-FET PET and PWI is based on different pathophysiological phenomena.

From K-space to Nucleotide: Insights Into the Radiogenomics of Brain Tumors

Radiogenomics is a relatively new and exciting field within radiology that links different imaging features with diverse genomic events. Genomics advances provided by the Cancer Genome Atlas and the Human Genome Project have enabled us to harness and integrate this information with noninvasive imaging phenotypes to create a better 3-dimensional understanding of tumor behavior and biology. Beyond imaging-histopathology, imaging genomic linkages provide an important layer of complexity that can help in evaluating and stratifying patients into clinical trials, monitoring treatment response, and enhancing patient outcomes. This article reviews some of the important radiogenomic literatures in brain tumors.