Differentiation between glioblastomas, solitary brain metastases, and primary cerebral lymphomas using diffusion tensor and dynamic susceptibility contrast-enhanced MR imaging

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 primary central nervous system lymphomas and gliomas according to pharmacokinetic parameters derived from dynamic contrast-enhanced magnetic resonance imaging

Purpose

It is difficult to differentiate between primary central nervous system lymphoma and primary glioblastoma due to their similar MRI findings. This study aimed to assess whether pharmacokinetic parameters derived from dynamic contrast-enhanced MRI could provide valuable insights for differentiation.

Methods

Seventeen cases of primary central nervous system lymphoma and twenty-one cases of glioblastoma as confirmed by pathology, were retrospectively analyzed. Pharmacokinetic parameters, including Ktrans, Kep, Ve, and the initial area under the Gd concentration curve, were measured from the enhancing tumor parenchyma, peritumoral parenchyma, and contralateral normal parenchyma. Statistical comparisons were made using Mann-Whitney U tests for Ve and Matrix Metallopeptidase-2, while independent samples t-tests were used to compare pharmacokinetic parameters in the mentioned regions and pathological indicators of enhancing tumor parenchyma, such as vascular endothelial growth factor and microvessel density. The pharmacokinetic parameters with statistical differences were evaluated using receiver-operating characteristics analysis. Except for the Wilcoxon rank sum test for Ve, the pharmacokinetic parameters were compared within the enhancing tumor parenchyma, peritumoral parenchyma, and contralateral normal parenchyma of the primary central nervous system lymphomas and glioblastomas using variance analysis and the least-significant difference method.

Results

Statistical differences were observed in Ktrans and Kep within the enhancing tumor parenchyma and in Kep within the peritumoral parenchyma between these two tumor types. Differences were also found in Matrix Metallopeptidase-2, vascular endothelial growth factor, and microvessel density within the enhancing tumor parenchyma of these tumors. When compared with the contralateral normal parenchyma, pharmacokinetic parameters within the peritumoral parenchyma and enhancing tumor parenchyma exhibited variations in glioblastoma and primary central nervous system lymphoma, respectively. Moreover, the receiver-operating characteristics analysis showed that the diagnostic efficiency of Kep in the peritumoral parenchyma was notably higher.

Conclusion

Pharmacokinetic parameters derived from dynamic contrast-enhanced MRI can differentiate primary central nervous system lymphoma and glioblastoma, especially Kep in the peritumoral parenchyma.

Predictors for the Differentiation between Glioblastoma, Primary Central Nervous System Lymphoma, and Metastasis in Patients with a Solitary Enhancing Intracranial Mass

Introduction  Differentiation between glioblastoma (GBM), primary central nervous system lymphoma (PCNSL), and metastasis is important in decision-making before surgery. However, these malignant brain tumors have overlapping features. This study aimed to identify predictors differentiating between GBM, PCNSL, and metastasis. Materials and Methods  Patients with a solitary intracranial enhancing tumor and a histopathological diagnosis of GBM, PCNSL, or metastasis were investigated. All patients with intracranial lymphoma had PCNSL without extracranial involvement. Demographic, clinical, and radiographic data were analyzed to determine their associations with the tumor types. Results  The predictors associated with GBM were functional impairment ( p  = 0.001), large tumor size ( p  < 0.001), irregular tumor margin ( p  < 0.001), heterogeneous contrast enhancement ( p  < 0.001), central necrosis ( p  < 0.001), intratumoral hemorrhage ( p  = 0.018), abnormal flow void ( p  < 0.001), and hypodensity component on noncontrast cranial computed tomography (CT) scan ( p  < 0.001). The predictors associated with PCNSL comprised functional impairment ( p  = 0.005), deep-seated tumor location ( p  = 0.006), homogeneous contrast enhancement ( p  < 0.001), absence of cystic appearance ( p  = 0.008), presence of hypointensity component on precontrast cranial T1-weighted magnetic resonance imaging (MRI; p  = 0.027), and presence of isodensity component on noncontrast cranial CT ( p  < 0.008). Finally, the predictors for metastasis were an infratentorial ( p  < 0.001) or extra-axial tumor location ( p  = 0.035), smooth tumor margin ( p  < 0.001), and presence of isointensity component on cranial fluid-attenuated inversion recovery MRI ( p  = 0.047). Conclusion  These predictors may be used to differentiate between GBM, PCNSL, and metastasis, and they are useful in clinical management.

Primary brain lymphoma and glioblastoma: evaluation of DCE T1 and DSC T2 MRI perfusion findings

BACKGROUND

The accurate differentiation of primary central nervous system lymphoma (PCNSL) from glioblastoma multiforme (GBM) is clinically crucial due to the different treatment strategies between them.

PURPOSE

To define magnetic resonance imaging (MRI) perfusion findings in PCNSL to make a safe distinction from GBM with dynamic contrast-enhanced (DCE) T1 and DSC T2 MRI perfusion findings.

MATERIAL AND METHODS

This retrospective analysis included 19 patients with histopathologically diagnosed PCNSL and 21 individuals with GBM. DCE T1 vascular permeability perfusion values including K-trans, Ve, Kep, IAUGC, and DSC T2 perfusion values including cerebral blood volume (CBV) and cerebral blood flow (CBF) in axial sections from the pathological lesion and contralateral normal brain parenchyma were measured quantitatively using region of interest analysis.

RESULTS

The study observed no statistically significant difference between patients with PCNSL (T/B cell) and GBM in the median values of DCE T1 perfusion ratios (P > 0.05). Nevertheless, the DSC T2 perfusion ratios showed a substantial distinction between the two groups. In contrast to patients with PCNSL (1.185 vs. 1.224, respectively), those with GBM had higher median levels of r-CBV and r-CBF (2.898 vs. 2.467, respectively; P 0.01). A cutoff value of ≤1.473 for r-CBV (Lesion/N) and ≤1.6005 for r-CBF (Lesion/N) was found to estimate the positivity of PCNSL.

CONCLUSION

DSC T2 MRI perfusion values showed lower r-CBV and r-CBF values in PCNSL patients compared to GBM patients. According to the findings, r-CBV and r-CBF are the most accurate MRI perfusion parameters for distinguishing between PCSNL and GBM.

An accessible deep learning tool for voxel-wise classification of brain malignancies from perfusion MRI

Noninvasive differential diagnosis of brain tumors is currently based on the assessment of magnetic resonance imaging (MRI) coupled with dynamic susceptibility contrast (DSC). However, a definitive diagnosis often requires neurosurgical interventions that compromise patients' quality of life. We apply deep learning on DSC images from histology-confirmed patients with glioblastoma, metastasis, or lymphoma. The convolutional neural network trained on ∼50,000 voxels from 40 patients provides intratumor probability maps that yield clinical-grade diagnosis. Performance is tested in 400 additional cases and an external validation cohort of 128 patients. The tool reaches a three-way accuracy of 0.78, superior to the conventional MRI metrics cerebral blood volume (0.55) and percentage of signal recovery (0.59), showing high value as a support diagnostic tool. Our open-access software, Diagnosis In Susceptibility Contrast Enhancing Regions for Neuro-oncology (DISCERN), demonstrates its potential in aiding medical decisions for brain tumor diagnosis using standard-of-care MRI.

Multiparametric analysis from dynamic susceptibility contrast-enhanced perfusion MRI to evaluate malignant brain tumors

BACKGROUND AND PURPOSE

Dynamic susceptibility contrast-enhanced (DSC) MR perfusion is a valuable technique for distinguishing brain tumors. Diagnostic potential of measurable parameters derived from preload leakage-corrected-DSC-MRI remains somewhat underexplored. This study aimed to evaluate these parameters for differentiating primary CNS lymphoma (PCNSL), glioblastoma, and metastasis.

METHODS

Thirty-nine patients with pathologically proven PCNSL (n = 14), glioblastoma (n = 14), and metastasis (n = 11) were analyzed. Five DSC parameters-relative CBV (rCBV), percentage of signal recovery (PSR), downward slope (DS), upward slope (US), and first-pass slope ratio-were derived from tumor-enhancing areas. Diagnostic performance was assessed using receiver operating characteristic curve analysis.

RESULTS

RCBV was higher in metastasis (4.58; interquartile range [IQR]: 2.54) and glioblastoma (3.98; IQR: 1.87), compared with PCNSL (1.46; IQR: 0.29; p = .00006 for both). rCBV better distinguished metastasis and glioblastoma from PCNSL, with an area under the curve (AUC) of 0.97 and 0.99, respectively. PSR was higher in PCNSL (88.11; IQR: 21.21) than metastases (58.30; IQR: 22.28; p = .0002), while glioblastoma (74.54; IQR: 21.23) presented almost significant trend-level differences compared to the others (p≈.05). AUCs were 0.79 (PCNSL vs. glioblastoma), 0.91 (PCNSL vs. metastasis), and 0.78 (glioblastoma vs. metastasis). DS and US parameters were statistically significant between glioblastoma (-109.92; IQR: 152.71 and 59.06; IQR: 52.87) and PCNSL (-47.36; IQR: 44.30 and 21.68; IQR: 16.85), presenting AUCs of 0.86 and 0.87.

CONCLUSION

Metastasis and glioblastoma can be better differentiated from PCNSL through rCBV. PSR demonstrated higher differential performance compared to the other parameters and seemed useful, allowing a proper distinction among all, particularly between metastasis and glioblastoma, where rCBV failed. Finally, DS and US were only helpful in differentiating glioblastoma from PCNSL.

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.

Differentiating primary central nervous system lymphoma from glioblastoma by time-dependent diffusion using oscillating gradient

BACKGROUND

This study aimed to elucidate the impact of effective diffusion time setting on apparent diffusion coefficient (ADC)-based differentiation between primary central nervous system lymphomas (PCNSLs) and glioblastomas (GBMs) and to investigate the usage of time-dependent diffusion magnetic resonance imaging (MRI) parameters.

METHODS

A retrospective study was conducted involving 21 patients with PCNSLs and 66 patients with GBMs using diffusion weighted imaging (DWI) sequences with oscillating gradient spin-echo (Δeff = 7.1 ms) and conventional pulsed gradient (Δeff = 44.5 ms). In addition to ADC maps at the two diffusion times (ADC7.1 ms and ADC44.5 ms), we generated maps of the ADC changes (cADC) and the relative ADC changes (rcADC) between the two diffusion times. Regions of interest were placed on enhancing regions and non-enhancing peritumoral regions. The mean and the fifth and 95th percentile values of each parameter were compared between PCNSLs and GBMs. The area under the receiver operating characteristic curve (AUC) values were used to compare the discriminating performances among the indices.

RESULTS

In enhancing regions, the mean and fifth and 95th percentile values of ADC44.5 ms and ADC7.1 ms in PCNSLs were significantly lower than those in GBMs (p = 0.02 for 95th percentile of ADC44.5 ms, p = 0.04 for ADC7.1 ms, and p < 0.01 for others). Furthermore, the mean and fifth and 95th percentile values of cADC and rcADC were significantly higher in PCNSLs than in GBMs (each p < 0.01). The AUC of the best-performing index for ADC7.1 ms was significantly lower than that for ADC44.5 ms (p < 0.001). The mean rcADC showed the highest discriminating performance (AUC = 0.920) among all indices. In peritumoral regions, no significant difference in any of the three indices of ADC44.5 ms, ADC7.1 ms, cADC, and rcADC was observed between PCNSLs and GBMs.

CONCLUSIONS

Effective diffusion time setting can have a crucial impact on the performance of ADC in differentiating between PCNSLs and GBMs. The time-dependent diffusion MRI parameters may be useful in the differentiation of these lesions.

Diffusion tensor imaging-based machine learning for IDH wild-type glioblastoma stratification to reveal the biological underpinning of radiomic features

INTRODUCTION

This study addresses the lack of systematic investigation into the prognostic value of hand-crafted radiomic features derived from diffusion tensor imaging (DTI) in isocitrate dehydrogenase (IDH) wild-type glioblastoma (GBM), as well as the limited understanding of the biological interpretation of individual DTI radiomic features and metrics.

AIMS

To develop and validate a DTI-based radiomic model for predicting prognosis in patients with IDH wild-type GBM and reveal the biological underpinning of individual DTI radiomic features and metrics.

RESULTS

The DTI-based radiomic signature was an independent prognostic factor (p < 0.001). Incorporating the radiomic signature into a clinical model resulted in a radiomic-clinical nomogram that predicted survival better than either the radiomic model or clinical model alone, with a better calibration and classification accuracy. Four categories of pathways (synapse, proliferation, DNA damage response, and complex cellular functions) were significantly correlated with the DTI-based radiomic features and DTI metrics.

CONCLUSION

The prognostic radiomic features derived from DTI are driven by distinct pathways involved in synapse, proliferation, DNA damage response, and complex cellular functions of GBM.

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.

Validation of multiparametric MRI based prediction model in identification of pseudoprogression in glioblastomas

BACKGROUND

Accurate differentiation of pseudoprogression (PsP) from tumor progression (TP) in glioblastomas (GBMs) is essential for appropriate clinical management and prognostication of these patients. In the present study, we sought to validate the findings of our previously developed multiparametric MRI model in a new cohort of GBM patients treated with standard therapy in identifying PsP cases.

METHODS

Fifty-six GBM patients demonstrating enhancing lesions within 6 months after completion of concurrent chemo-radiotherapy (CCRT) underwent anatomical imaging, diffusion and perfusion MRI on a 3 T magnet. Subsequently, patients were classified as TP + mixed tumor (n = 37) and PsP (n = 19). When tumor specimens were available from repeat surgery, histopathologic findings were used to identify TP + mixed tumor (> 25% malignant features; n = 34) or PsP (< 25% malignant features; n = 16). In case of non-availability of tumor specimens, ≥ 2 consecutive conventional MRIs using mRANO criteria were used to determine TP + mixed tumor (n = 3) or PsP (n = 3). The multiparametric MRI-based prediction model consisted of predictive probabilities (PP) of tumor progression computed from diffusion and perfusion MRI derived parameters from contrast enhancing regions. In the next step, PP values were used to characterize each lesion as PsP or TP+ mixed tumor. The lesions were considered as PsP if the PP value was < 50% and TP+ mixed tumor if the PP value was ≥ 50%. Pearson test was used to determine the concordance correlation coefficient between PP values and histopathology/mRANO criteria. The area under ROC curve (AUC) was used as a quantitative measure for assessing the discriminatory accuracy of the prediction model in identifying PsP and TP+ mixed tumor.

RESULTS

Multiparametric MRI model correctly predicted PsP in 95% (18/19) and TP+ mixed tumor in 57% of cases (21/37) with an overall concordance rate of 70% (39/56) with final diagnosis as determined by histopathology/mRANO criteria. There was a significant concordant correlation coefficient between PP values and histopathology/mRANO criteria (r = 0.56; p < 0.001). The ROC analyses revealed an accuracy of 75.7% in distinguishing PsP from TP+ mixed tumor. Leave-one-out cross-validation test revealed that 73.2% of cases were correctly classified as PsP and TP + mixed tumor.

CONCLUSIONS

Our multiparametric MRI based prediction model may be helpful in identifying PsP in GBM patients.

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.

Differentiation of Perilesional Edema in Glioblastomas and Brain Metastases: Comparison of Diffusion Tensor Imaging, Neurite Orientation Dispersion and Density Imaging and Diffusion Microstructure Imaging

Although the free water content within the perilesional T2 hyperintense region should differ between glioblastomas (GBM) and brain metastases based on histological differences, the application of classical MR diffusion models has led to inconsistent results regarding the differentiation between these two entities. Whereas diffusion tensor imaging (DTI) considers the voxel as a single compartment, multicompartment approaches such as neurite orientation dispersion and density imaging (NODDI) or the recently introduced diffusion microstructure imaging (DMI) allow for the calculation of the relative proportions of intra- and extra-axonal and also free water compartments in brain tissue. We investigate the potential of water-sensitive DTI, NODDI and DMI metrics to detect differences in free water content of the perilesional T2 hyperintense area between histopathologically confirmed GBM and brain metastases. Respective diffusion metrics most susceptible to alterations in the free water content (MD, V-ISO, V-CSF) were extracted from T2 hyperintense perilesional areas, normalized and compared in 24 patients with GBM and 25 with brain metastases. DTI MD was significantly increased in metastases (p = 0.006) compared to GBM, which was corroborated by an increased DMI V-CSF (p = 0.001), while the NODDI-derived ISO-VF showed only trend level increase in metastases not reaching significance (p = 0.060). In conclusion, diffusion MRI metrics are able to detect subtle differences in the free water content of perilesional T2 hyperintense areas in GBM and metastases, whereas DMI seems to be superior to DTI and NODDI.

Prognostication of overall survival in patients with brain metastases using diffusion tensor imaging and dynamic susceptibility contrast-enhanced MRI

OBJECTIVES

To investigate the prognostic utility of DTI and DSC-PWI perfusion-derived parameters in brain metastases patients.

METHODS

Retrospective analyses of DTI-derived parameters (MD, FA, CL, CP, and CS) and DSC-perfusion PWI-derived rCBVmax from 101 patients diagnosed with brain metastases prior to treatment were performed. Using semi-automated segmentation, DTI metrics and rCBVmax were quantified from enhancing areas of the dominant metastatic lesion. For each metric, patients were classified as short- and long-term survivors based on analysis of the best coefficient for each parameter and percentile to separate the groups. Kaplan-Meier analysis was used to compare mOS between these groups. Multivariate survival analysis was subsequently conducted. A correlative histopathologic analysis was performed in a subcohort (n = 10) with DTI metrics and rCBVmax on opposite ends of the spectrum.

RESULTS

Significant differences in mOS were observed for MDmin (p < 0.05), FA (p < 0.01), CL (p < 0.05), and CP (p < 0.01) and trend toward significance for rCBVmax (p = 0.07) between the two risk groups, in the univariate analysis. On multivariate analysis, the best predictive survival model was comprised of MDmin (p = 0.05), rCBVmax (p < 0.05), RPA (p < 0.0001), and number of lesions (p = 0.07). On histopathology, metastatic tumors showed significant differences in the amount of stroma depending on the combination of DTI metrics and rCBVmax values. Patients with high stromal content demonstrated poorer mOS.

CONCLUSION

Pretreatment DTI-derived parameters, notably MDmin and rCBVmax, are promising imaging markers for prognostication of OS in patients with brain metastases. Stromal cellularity may be a contributing factor to these differences.

ADVANCES IN KNOWLEDGE

The correlation of DTI-derived metrics and perfusion MRI with patient outcomes has not been investigated in patients with treatment naïve brain metastasis. DTI and DSC-PWI can aid in therapeutic decision-making by providing additional clinical guidance.

Fully automated radiomics-based machine learning models for multiclass classification of single brain tumors: Glioblastoma, lymphoma, and metastasis

BACKGROUND AND PURPOSE

To investigate the diagnostic performance of fully automated radiomics-based models for multiclass classification of a single enhancing brain tumor among glioblastoma, central nervous system lymphoma, and metastasis.

MATERIALS AND METHODS

The training and test sets were comprised of 538 cases (300 glioblastomas, 73 lymphomas, and 165 metastases) and 169 cases (101 glioblastomas, 29 lymphomas, and 39 metastases), respectively. After fully automated segmentation, radiomic features were extracted. Three conventional machine learning classifiers, including least absolute shrinkage and selection operator (LASSO), adaptive boosting (Adaboost), and support vector machine with the linear kernel (SVC), combined with one of four feature selection methods, including forward sequential feature selection, F score, mutual information, and LASSO, were trained. Additionally, one ensemble classifier based on the three classifiers was used. The diagnostic performance of the optimized models was tested in the test set using the accuracy, F1-macro score, and the area under the receiver operating characteristic curve (AUCROC).

RESULTS

The best performance was achieved when the LASSO was used as a feature selection method. In the test set, the best performance was achieved by the ensemble classifier, showing an accuracy of 76.3% (95% CI, 70.0-82.7), a F1-macro score of 0.704, and an AUCROC of 0.878.

CONCLUSION

Our fully automated radiomics-based models for multiclass classification might be useful for differential diagnosis of a single enhancing brain tumor with a good diagnostic performance and generalizability.

Diffuse Large B-Cell Epstein-Barr Virus-Positive Primary CNS Lymphoma in Non-AIDS Patients: High Diagnostic Accuracy of DSC Perfusion Metrics

BACKGROUND AND PURPOSE

Immunodeficiency-associated CNS lymphoma may occur in different clinical scenarios beyond AIDS. This subtype of CNS lymphoma is diffuse large B-cell and Epstein-Barr virus-positive. Its accurate presurgical diagnosis is often unfeasible because it appears as ring-enhancing lesions mimicking glioblastoma or metastasis. In this article, we describe clinicoradiologic features and test the performance of DSC-PWI metrics for presurgical identification.

MATERIALS AND METHODS

Patients without AIDS with histologically confirmed diffuse large B-cell Epstein-Barr virus-positive primary CNS lymphoma (December 2010 to January 2022) and diagnostic MR imaging without onco-specific treatment were retrospectively studied. Clinical, demographic, and conventional imaging data were reviewed. Previously published DSC-PWI time-intensity curve analysis methodology, to presurgically identify primary CNS lymphoma, was used in this particular lymphoma subtype and compared with a prior cohort of 33 patients with Epstein-Barr virus-negative CNS lymphoma, 35 with glioblastoma, and 36 with metastasis data. Normalized curves were analyzed and compared on a point-by-point basis, and previously published classifiers were tested. The standard percentage of signal recovery and CBV values were also evaluated.

RESULTS

Seven patients with Epstein-Barr virus-positive primary CNS lymphoma were included in the study. DSC-PWI normalized time-intensity curve analysis performed the best for presurgical identification of Epstein-Barr virus-positive CNS lymphoma (area under the receiver operating characteristic curve of 0.984 for glioblastoma and 0.898 for metastasis), followed by the percentage of signal recovery (0.833 and 0.873) and CBV (0.855 and 0.687).

CONCLUSIONS

When a necrotic tumor is found in a potentially immunocompromised host, neuroradiologists should consider Epstein-Barr virus-positive CNS lymphoma. DSC-PWI could be very useful for presurgical characterization, with especially strong performance of normalized time-intensity curves.

Differentiation of Intracerebral Tumor Entities with Quantitative Contrast Attenuation and Iodine Mapping in Dual-Layer Computed Tomography

Purpose: To investigate if quantitative contrast enhancement and iodine mapping of common brain tumor (BT) entities may correctly differentiate between tumor etiologies in standardized stereotactic CT protocols. Material and Methods: A retrospective monocentric study of 139 consecutive standardized dual-layer dual-energy CT (dlDECT) scans conducted prior to the stereotactic needle biopsy of untreated primary brain tumor lesions. Attenuation of contrast-enhancing BT was derived from polyenergetic images as well as spectral iodine density maps (IDM) and their contrast-to-noise-ratios (CNR) were determined using ROI measures in contrast-enhancing BT and healthy contralateral white matter. The measures were correlated to histopathology regarding tumor entity, isocitrate dehydrogenase (IDH) and MGMT mutation status. Results: The cohort included 52 female and 76 male patients, mean age of 59.4 (±17.1) years. Brain lymphomas showed the highest attenuation (IDM CNR 3.28 ± 1,23), significantly higher than glioblastoma (2.37 ± 1.55, p < 0.005) and metastases (1.95 ± 1.14, p < 0.02), while the differences between glioblastomas and metastases were not significant. These strongly enhancing lesions differed from oligodendroglioma and astrocytoma (Grade II and III) that showed IDM CNR in the range of 1.22−1.27 (±0.45−0.82). Conventional attenuation measurements in DLCT data performed equally or slightly superior to iodine density measurements. Conclusion: Quantitative attenuation and iodine density measurements of contrast-enhancing brain tumors are feasible imaging biomarkers for the discrimination of cerebral tumor lesions but not specifically for single tumor entities. CNR based on simple HU measurements performed equally or slightly superior to iodine quantification.

Revisiting the use of proton magnetic resonance spectroscopy in distinguishing between primary and secondary malignant tumors of the central nervous system

BACKGROUND AND PURPOSE

Conventional magnetic resonance images (MRI) has limitations in distinguishing primary from secondary brain tumors. Proton magnetic resonance spectroscopy (1H-MRS) allows evaluation of the concentration of metabolites in a brain lesion and, hence, better characterization of the tumor. Considering that an accurate diagnosis determines the choice of treatment, our purpose was to assess the usefulness of spectroscopy data for differentiating between primary and secondary brain neoplasms.

MATERIALS AND METHODS

We undertook a retrospective analysis of 61 MRI and 1H-MRS images of patients with histologically confirmed tumors (30 primary tumors and 31 metastatic tumors). The metabolite ratios of Cho/Cr and NAA/Cr at short TE were determined from spectroscopic curves, with a single voxel positioned in the enhancing tumor. Additional variables analyzed along with the metabolites, like as age and gender, allowed the construction of a logistic regression model to predict the tumor's nature. The statistical analysis was done using the R software (version 4.0.3 R Core Team, 2020).

RESULTS

The mean NAA/Cr and Cho/Cr ratios were higher in secondary tumors, with a good correlation between NAA/Cr and Cho/Cr (r = 0.61). The mean age of patients with primary tumors was lower than for secondary tumors (43.9 vs 55.9, respectively). Receiver operating characteristic analysis yielded a cut-off value of 0.4 for the NAA/Cr ratio with an accuracy of 73.8%, a sensitivity of 73.3% and a specificity of 74.2% in predicting metastatic tumors.

CONCLUSION

The model was reasonable in predicting the nature of the tumor and provides an additional tool for analyzing brain tumors.

Towards updated understanding of brain metastasis

Brain metastasis (BM) is a common complication in cancer patients with advanced disease and attributes to treatment failure and final mortality. Currently there are several therapeutic options available; however these are only suitable for limited subpopulation: surgical resection or radiosurgery for cases with a limited number of lesions, targeted therapies for approximately 18% of patients, and immune checkpoint inhibitors with a response rate of 20-30%. Thus, there is a pressing need for development of novel diagnostic and therapeutic options. This overview article aims to provide research advances in disease model, targeted therapy, blood brain barrier (BBB) opening strategies, imaging and its incorporation with artificial intelligence, external radiotherapy, and internal targeted radionuclide theragnostics. Finally, a distinct type of BM, leptomeningeal metastasis is also covered.

Metabolic and physiologic magnetic resonance imaging in distinguishing true progression from pseudoprogression in patients with glioblastoma

Pseudoprogression (PsP) refers to treatment-related clinico-radiologic changes mimicking true progression (TP) that occurs in patients with glioblastoma (GBM), predominantly within the first 6 months after the completion of surgery and concurrent chemoradiation therapy (CCRT) with temozolomide. Accurate differentiation of TP from PsP is essential for making informed decisions on appropriate therapeutic intervention as well as for prognostication of these patients. Conventional neuroimaging findings are often equivocal in distinguishing between TP and PsP and present a considerable diagnostic dilemma to oncologists and radiologists. These challenges have emphasized the need for developing alternative imaging techniques that may aid in the accurate diagnosis of TP and PsP. In this review, we encapsulate the current state of knowledge in the clinical applications of commonly used metabolic and physiologic magnetic resonance (MR) imaging techniques such as diffusion and perfusion imaging and proton spectroscopy in distinguishing TP from PsP. We also showcase the potential of promising imaging techniques, such as amide proton transfer and amino acid-based positron emission tomography, in providing useful information about the treatment response. Additionally, we highlight the role of "radiomics", which is an emerging field of radiology that has the potential to change the way in which advanced MR techniques are utilized in assessing treatment response in GBM patients. Finally, we present our institutional experiences and discuss future perspectives on the role of multiparametric MR imaging in identifying PsP in GBM patients treated with "standard-of-care" CCRT as well as novel/targeted therapies.

Machine Learning in Differentiating Gliomas from Primary CNS Lymphomas: A Systematic Review, Reporting Quality, and Risk of Bias Assessment

BACKGROUND

Differentiating gliomas and primary CNS lymphoma represents a diagnostic challenge with important therapeutic ramifications. Biopsy is the preferred method of diagnosis, while MR imaging in conjunction with machine learning has shown promising results in differentiating these tumors.

PURPOSE

Our aim was to evaluate the quality of reporting and risk of bias, assess data bases with which the machine learning classification algorithms were developed, the algorithms themselves, and their performance.

DATA SOURCES

Ovid EMBASE, Ovid MEDLINE, Cochrane Central Register of Controlled Trials, and the Web of Science Core Collection were searched according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

STUDY SELECTION

From 11,727 studies, 23 peer-reviewed studies used machine learning to differentiate primary CNS lymphoma from gliomas in 2276 patients.

DATA ANALYSIS

Characteristics of data sets and machine learning algorithms were extracted. A meta-analysis on a subset of studies was performed. Reporting quality and risk of bias were assessed using the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) and Prediction Model Study Risk Of Bias Assessment Tool.

DATA SYNTHESIS

The highest area under the receiver operating characteristic curve (0.961) and accuracy (91.2%) in external validation were achieved by logistic regression and support vector machines models using conventional radiomic features. Meta-analysis of machine learning classifiers using these features yielded a mean area under the receiver operating characteristic curve of 0.944 (95% CI, 0.898-0.99). The median TRIPOD score was 51.7%. The risk of bias was high for 16 studies.

LIMITATIONS

Exclusion of abstracts decreased the sensitivity in evaluating all published studies. Meta-analysis had high heterogeneity.

CONCLUSIONS

Machine learning-based methods of differentiating primary CNS lymphoma from gliomas have shown great potential, but most studies lack large, balanced data sets and external validation. Assessment of the studies identified multiple deficiencies in reporting quality and risk of bias. These factors reduce the generalizability and reproducibility of the findings.

Advancements in Oncology with Artificial Intelligence—A Review Article

Simple Summary

With the advancement of artificial intelligence, including machine learning, the field of oncology has seen promising results in cancer detection and classification, epigenetics, drug discovery, and prognostication. In this review, we describe what artificial intelligence is and its function, as well as comprehensively summarize its evolution and role in breast, colorectal, and central nervous system cancers. Understanding the origin and current accomplishments might be essential to improve the quality, accuracy, generalizability, cost-effectiveness, and reliability of artificial intelligence models that can be used in worldwide clinical practice. Students and researchers in the medical field will benefit from a deeper understanding of how to use integrative AI in oncology for innovation and research.

Abstract

Well-trained machine learning (ML) and artificial intelligence (AI) systems can provide clinicians with therapeutic assistance, potentially increasing efficiency and improving efficacy. ML has demonstrated high accuracy in oncology-related diagnostic imaging, including screening mammography interpretation, colon polyp detection, glioma classification, and grading. By utilizing ML techniques, the manual steps of detecting and segmenting lesions are greatly reduced. ML-based tumor imaging analysis is independent of the experience level of evaluating physicians, and the results are expected to be more standardized and accurate. One of the biggest challenges is its generalizability worldwide. The current detection and screening methods for colon polyps and breast cancer have a vast amount of data, so they are ideal areas for studying the global standardization of artificial intelligence. Central nervous system cancers are rare and have poor prognoses based on current management standards. ML offers the prospect of unraveling undiscovered features from routinely acquired neuroimaging for improving treatment planning, prognostication, monitoring, and response assessment of CNS tumors such as gliomas. By studying AI in such rare cancer types, standard management methods may be improved by augmenting personalized/precision medicine. This review aims to provide clinicians and medical researchers with a basic understanding of how ML works and its role in oncology, especially in breast cancer, colorectal cancer, and primary and metastatic brain cancer. Understanding AI basics, current achievements, and future challenges are crucial in advancing the use of AI in oncology.

A Deep Learning Model for Preoperative Differentiation of Glioblastoma, Brain Metastasis and Primary Central Nervous System Lymphoma: A Pilot Study

Background

Neuroimaging differentiation of glioblastoma, primary central nervous system lymphoma (PCNSL) and solitary brain metastasis (BM) remains challenging in specific cases showing similar appearances or atypical features. Overall, advanced MRI protocols have high diagnostic reliability, but their limited worldwide availability, coupled with the overlapping of specific neuroimaging features among tumor subgroups, represent significant drawbacks and entail disparities in the planning and management of these oncological patients.

Objective

To evaluate the classification performance metrics of a deep learning algorithm trained on T1-weighted gadolinium-enhanced (T1Gd) MRI scans of glioblastomas, atypical PCNSLs and BMs.

Materials and Methods

We enrolled 121 patients (glioblastoma: n=47; PCNSL: n=37; BM: n=37) who had undergone preoperative T1Gd-MRI and histopathological confirmation. Each lesion was segmented, and all ROIs were exported in a DICOM dataset. The patient cohort was then split in a training and hold-out test sets following a 70/30 ratio. A Resnet101 model, a deep neural network (DNN), was trained on the training set and validated on the hold-out test set to differentiate glioblastomas, PCNSLs and BMs on T1Gd-MRI scans.

Results

The DNN achieved optimal classification performance in distinguishing PCNSLs (AUC: 0.98; 95%CI: 0.95 - 1.00) and glioblastomas (AUC: 0.90; 95%CI: 0.81 - 0.97) and moderate ability in differentiating BMs (AUC: 0.81; 95%CI: 0.70 - 0.95). This performance may allow clinicians to correctly identify patients eligible for lesion biopsy or surgical resection.

Conclusion

We trained and internally validated a deep learning model able to reliably differentiate ambiguous cases of PCNSLs, glioblastoma and BMs by means of T1Gd-MRI. The proposed predictive model may provide a low-cost, easily-accessible and high-speed decision-making support for eligibility to diagnostic brain biopsy or maximal tumor resection in atypical cases.

3D pseudo-continuous arterial spin labeling-MRI (3D PCASL-MRI) in the differential diagnosis between glioblastomas and primary central nervous system lymphomas

PURPOSE

The aim of the study was to compare the parameters of blood flow in glioblastomas and primary central nervous system lymphomas (PCNSLs), measured by pseudo-continuous arterial spin labeling MRI (3D PCASL), and to determine the informativeness of this method in the differential diagnosis between these lesions.

METHODS

The study included MRI data of 139 patients with PCNSL (n = 21) and glioblastomas (n = 118), performed in the Burdenko Neurosurgical Center. No patients received chemotherapy, hormone therapy, or radiation therapy prior to MRI. On the 3D PCASL perfusion map, the absolute and normalized values of tumor blood flow were calculated in the glioblastoma and PCNSL groups (maxTBFmean and nTBF).

RESULTS

MaxTBFmean and nTBF in the glioblastoma group were significantly higher than those in the PCNSL group: 168.9 ml/100 g/min versus 65.6 and 9.3 versus 3.7, respectively (p < 0.001). Arterial spin labeling perfusion had high sensitivity (86% for maxTBFmean, 95% for nTBF) and specificity (77% for maxTBFmean, 73% for nTBF) in the differential diagnosis between PCNSL and glioblastomas. Blood flow thresholds were 98.9 ml/100 g/min using absolute blood flow values and 6.1 using normalized values, AUC > 0.88.

CONCLUSION

The inclusion of 3D PCASL in the standard MRI protocol can increase the specificity of the differential diagnosis between glioblastomas and PCNSL.

Voxel-level analysis of normalized DSC-PWI time-intensity curves: a potential generalizable approach and its proof of concept in discriminating glioblastoma and metastasis

OBJECTIVE

Standard DSC-PWI analyses are based on concrete parameters and values, but an approach that contemplates all points in the time-intensity curves and all voxels in the region-of-interest may provide improved information, and more generalizable models. Therefore, a method of DSC-PWI analysis by means of normalized time-intensity curves point-by-point and voxel-by-voxel is constructed, and its feasibility and performance are tested in presurgical discrimination of glioblastoma and metastasis.

METHODS

In this retrospective study, patients with histologically confirmed glioblastoma or solitary-brain-metastases and presurgical-MR with DSC-PWI (August 2007-March 2020) were retrieved. The enhancing tumor and immediate peritumoral region were segmented on CE-T1wi and coregistered to DSC-PWI. Time-intensity curves of the segmentations were normalized to normal-appearing white matter. For each participant, average and all-voxel-matrix of normalized-curves were obtained. The 10 best discriminatory time-points between each type of tumor were selected. Then, an intensity-histogram analysis on each of these 10 time-points allowed the selection of the best discriminatory voxel-percentile for each. Separate classifier models were trained for enhancing tumor and peritumoral region using binary logistic regressions.

RESULTS

A total of 428 patients (321 glioblastomas, 107 metastases) fulfilled the inclusion criteria (256 men; mean age, 60 years; range, 20-86 years). Satisfactory results were obtained to segregate glioblastoma and metastases in training and test sets with AUCs 0.71-0.83, independent accuracies 65-79%, and combined accuracies up to 81-88%.

CONCLUSION

This proof-of-concept study presents a different perspective on brain MR DSC-PWI evaluation by the inclusion of all time-points of the curves and all voxels of segmentations to generate robust diagnostic models of special interest in heterogeneous diseases and populations. The method allows satisfactory presurgical segregation of glioblastoma and metastases.

KEY POINTS

• An original approach to brain MR DSC-PWI analysis, based on a point-by-point and voxel-by-voxel assessment of normalized time-intensity curves, is presented. • The method intends to extract optimized information from MR DSC-PWI sequences by impeding the potential loss of information that may represent the standard evaluation of single concrete perfusion parameters (cerebral blood volume, percentage of signal recovery, or peak height) and values (mean, maximum, or minimum). • The presented approach may be of special interest in technically heterogeneous samples, and intrinsically heterogeneous diseases. Its application enables satisfactory presurgical differentiation of GB and metastases, a usual but difficult diagnostic challenge for neuroradiologist with vital implications in patient management.

Differentiation Between Primary Central Nervous System Lymphoma and Atypical Glioblastoma Based on MRI Morphological Feature and Signal Intensity Ratio: A Retrospective Multicenter Study

Objectives

To investigate the value of morphological feature and signal intensity ratio (SIR) derived from conventional magnetic resonance imaging (MRI) in distinguishing primary central nervous system lymphoma (PCNSL) from atypical glioblastoma (aGBM).

Methods

Pathology-confirmed PCNSLs (n = 93) or aGBMs (n = 48) from three institutions were retrospectively enrolled and divided into training cohort (n = 98) and test cohort (n = 43). Morphological features and SIRs were compared between PCNSL and aGBM. Using linear discriminant analysis, multiple models were constructed with SIRs and morphological features alone or jointly, and the diagnostic performances were evaluated via receiver operating characteristic (ROC) analysis. Areas under the curves (AUCs) and accuracies (ACCs) of the models were compared with the radiologists’ assessment.

Results

Incision sign, T₂ pseudonecrosis sign, reef sign and peritumoral leukomalacia sign were associated with PCNSL (training and overall cohorts, P < 0.05). Increased T₁ ratio, decreased T₂ ratio and T₂/T₁ ratio were predictive of PCNSL (all P < 0.05). ROC analysis showed that combination of morphological features and SIRs achieved the best diagnostic performance for differentiation of PCNSL and aGBM with AUC/ACC of 0.899/0.929 for the training cohort, AUC/ACC of 0.794/0.837 for the test cohort and AUC/ACC of 0.869/0.901 for the overall cohort, respectively. Based on the overall cohort, two radiologists could distinguish PCNSL from aGBM with AUC/ACC of 0.732/0.724 for radiologist A and AUC/ACC of 0.811/0.829 for radiologist B.

Conclusion

MRI morphological features can help differentiate PCNSL from aGBM. When combined with SIRs, the diagnostic performance was better than that of radiologists’ assessment.

Diffusion tensor imaging derived metrics in high grade glioma and brain metastasis differentiation

Background: Pretreatment differentiation between glioblastoma and metastasis is a frequently encountered dilemma in neurosurgical practice. Distinction is required for precise planning of resection or radiotherapy, and also for defining further diagnostic procedures. Morphology and spectroscopy imaging features are not specific and frequently overlap. This limitation of magnetic resonance imaging and magnetic resonance spectroscopy was the reason to initiate this study. The aim of the present study was to determine whether the dataset of diffusion tensor imaging metrics contains information which may be used for the distinction between primary and secondary intra-axial neoplasms. Methods: Two diffusion tensor imaging parameters were measured in 81 patients with an expansive, ring-enhancing, intra-axial lesion on standard magnetic resonance imaging (1.5 T system). All tumors were histologically verified glioblastoma or secondary deposit. For qualitative analysis, two regions of interest were defined: intratumoral and immediate peritumoral region (locations 1 and 2, respectively). Fractional anisotropy and mean difusivity values of both groups were compared. Additional test was performed to determine if there was a significant difference in mean values between two locations. Results: A statistically significant difference was found in fractional anisotropy values among two locations, with decreasing values in the direction of neoplastic infiltration, although such difference was not observed in fractional anisotropy values in the group with secondary tumors. Mean difusivity values did not appear helpful in differentiation between these two entities. In both groups there was no significant difference in mean difusivity values, neither in intratumoral nor in peritumoral location. Conclusion: The results of our study justify associating the diffusion tensor imaging technique to conventional morphologic magnetic resonance imaging as an additional diagnostic tool for the distinction between primary and secondary intra-axial lesions. Quantitative analysis of diffusion tensor imaging metric, in particular measurement of fractional anisotropy in peritumoral edema facilitates accurate diagnosis.

The Application of Image Segmentation to Determine the Ratio of Peritumoral Edema Area to Tumor Area on Primary Malignant Brain Tumor and Metastases through Conventional Magnetic Resonance Imaging

BACKGROUND: Primary malignant brain tumor and metastases on the brain have a similar pattern in conventional Magnetic Resonance Imaging (MRI), even though both require entirely different treatment and management. The pathophysiological difference of peritumoral edema can help to distinguish the case of primary malignant brain tumor and brain metastases. AIM: This study aimed to analyze the ratio of the area of peritumoral edema to the tumor using Otsu’s method of image segmentation technique with a user-friendly Graphical User Interface (GUI). METODS: Data was prepared by obtaining the examination results of Anatomical Pathology and MRI imaging. The area of peritumoral edema was identified from MRI image segmentation with T2/FLAIR sequence. While the area of tumor was identified using MRI image segmentation with T1 sequence. RESULTS: The Mann-Whitney test was employed to analyze the ratio of the area of peritumoral edema to tumor on both groups. Data testing produced a significance level of 0.013 (p < 0.05) with a median value (Nmax-Nmin) of 1.14 (3.31-0.08) for the primary malignant brain tumor group and a median value (Nmax-Nmin) of 1.17 (10.30-0.90) for the brain metastases group. CONCLUSIONS: There was a significant difference in the ratio of the area of peritumoral edema to the area of tumor from both groups, in which brain metastases have a greater value than the primary malignant brain tumor.

Histogram analysis based on diffusion kurtosis imaging: Differentiating glioblastoma multiforme from single brain metastasis and comparing the diagnostic performance of two region of interest placements

PURPOSE

To assess the value of histogram analysis, using diffusion kurtosis imaging (DKI), in differentiating glioblastoma multiforme (GBM) from single brain metastasis (SBM) and to compare the diagnostic efficiency of different region of interest (ROI) placements.

METHOD

Sixty-seven patients with histologically confirmed GBM (n = 35) and SBM (n = 32) were recruited. Two ROIs-the contrast-enhanced area and whole-tumor area-were delineated across all slices. Eleven histogram parameters of fractional anisotropy (FA), mean diffusivity (MD), and mean kurtosis (MK) from both ROIs were calculated. All histogram parameter values were compared between GBM and SBM, using the Mann-Whitney U test. The accuracies of different histogram parameters were compared using the McNemar test. Receiver operating characteristic (ROC) analyses were conducted to assess the diagnostic performance.

RESULTS

In the contrast-enhanced area, FA₁₀, FA₂₅, FA₇₅, FA₉₀, FA_mean, FA_median, FA_max, MD_max, MD_skewness, and MK_skewness were significantly higher for GBM than for SBM. FA_skewness was significantly lower for GBM than for SBM. FA₂₅ (0.815) had the highest area under the curve (AUC). In the whole-tumor area, FA₁₀, FA₂₅, FA₇₅, FA₉₀, FA_SD, FA_mean, FA_median, FA_max, MD_max, MD_skewness, and MK_skewness were significantly higher for GBM than for SBM. FA_median (0.805) had the highest AUC. The accuracy of FA₂₅ in the contrast-enhanced area was significantly higher than that of the FA_median in the whole-tumor area.

CONCLUSIONS

GBM and SBM can be differentiated using the DKI-based histogram analysis. Placing the ROI on the contrast-enhanced area results in better discrimination.

Neural Fiber Integrity in High- Versus Low-Grade Glioma using Probabilistic Fiber Tracking

RATIONALE AND OBJECTIVES

Gliomatous tumors are known to affect neural fiber integrity, either by displacement or destruction. The aim of this study is to investigate the integrity and distribution of the white matter tracts within and around the glioma regions using probabilistic fiber tracking.

MATERIAL AND METHODS

Forty-two glioma patients were subjected to MRI using a standard tumor protocol with diffusion tensor imaging (DTI). The tumor and peritumor regions were delineated using snake model with reference to structural and diffusion MRI. A preprocessing pipeline of the structural MRI image, DTI data, and tumor regions was implemented. Tractography was performed to delineate the white matter (WM) tracts in the selected tumor regions via probabilistic fiber tracking. DTI indices were investigated through comparative mapping of WM tracts and tumor regions in low-grade gliomas (LGG) and high-grade gliomas (HGG).

RESULTS

Significant differences were seen in the planar tensor (C_p) in peritumor regions; mean diffusivity, axial diffusivity and pure isotropic diffusion in solid-enhancing tumor regions; and fractional anisotropy, axial diffusivity, pure anisotropic diffusion (q), total magnitude of diffusion tensor (L), relative anisotropy, C_p and spherical tensor (C_s) in solid nonenhancing tumor regions for affected WM tracts. In most cases of HGG, the WM tracts were not completely destroyed, but found intact inside the tumor.

DISCUSSION

Probabilistic fiber tracking revealed the existence and distribution of WM tracts inside tumor core for both LGG and HGG groups. There were more DTI indices in the solid nonenhancing tumor region, which showed significant differences between LGG and HGG.

Advanced MRI assessment of non-enhancing peritumoral signal abnormality in brain lesions

Evaluation of Central Nervous System (CNS) focal lesions has been classically made focusing on the assessment solid or enhancing component. However, the assessment of solitary peripherally enhancing lesions where the differential diagnosis includes High-Grade Gliomas (HGG) and metastasis, is usually challenging. Several studies have tried to address the characteristics of peritumoral non-enhancing areas, for better characterization of these lesions. Peritumoral hyperintense T2/FLAIR signal abnormality predominantly contains infiltrating tumor cells in HGG whereas CNS metastasis induce pure vasogenic edema. In addition, the accurate determination of the real extension of HGG is critical for treatment selection and outcome. Conventional MRI sequences are limited in distinguishing infiltrating neoplasm from vasogenic edema. Advanced MRI sequences like Diffusion Weighted Imaging (DWI), Diffusion Tensor Imaging (DTI), Perfusion Weighted Imaging (PWI) and MR spectroscopy (MRS) have all been utilized for this aim with acceptable results. Other advanced MRI approaches, less explored for this task such as Arterial Spin Labelling (ASL), Diffusion Kurtosis Imaging (DKI), T2 relaxometry or Amide Proton Transfer (APT) are also showning promising results in this scenario. In this article, we will discuss the physiopathological basis of peritumoral T2/FLAIR signal abnormality and review potential applications of advanced MRI sequences for its evaluation.

Differentiating Glioblastomas from Solitary Brain Metastases: An Update on the Current Literature of Advanced Imaging Modalities

Differentiating between glioblastomas and solitary brain metastases proves to be a challenging diagnosis for neuroradiologists, as both present with imaging patterns consisting of peritumoral hyperintensities with similar intratumoral texture on traditional magnetic resonance imaging sequences. Early diagnosis is paramount, as each pathology has completely different methods of clinical assessment. In the past decade, recent developments in advanced imaging modalities enabled providers to acquire a more accurate diagnosis earlier in the patient's clinical assessment, thus optimizing clinical outcome. Dynamic susceptibility contrast has been optimized for detecting relative cerebral blood flow and relative cerebral blood volume. Diffusion tensor imaging can be used to detect changes in mean diffusivity. Neurite orientation dispersion and density imaging is an innovative modality detecting changes in intracellular volume fraction, isotropic volume fraction, and extracellular volume fraction. Magnetic resonance spectroscopy is able to assist by providing a metabolic descriptor while detecting variable ratios of choline/N-acetylaspartate, choline/creatine, and N-acetylaspartate/creatine. Finally, radiomics and machine learning algorithms have been devised to assist in improving diagnostic accuracy while often utilizing more than one advanced imaging protocol per patient. In this review, we provide an update on all the current evidence regarding the identification and differentiation of glioblastomas from solitary brain metastases.

Phosphorous Magnetic Resonance Spectroscopy to Detect Regional Differences of Energy and Membrane Metabolism in Naïve Glioblastoma Multiforme

Simple Summary

Glioblastoma multiforme is a highly aggressive brain tumor, tending to infiltrate even larger zones of brain tissue than visible on conventional magnetic resonance imaging. By application of phosphorus magnetic resonance spectroscopy in patients with naïve glioblastoma multiforme, we tried to demonstrate changes in energy and membrane metabolism not only in affected regions but also in distant brain regions, the opposite brain hemisphere, and in comparison to healthy volunteers. We found reduced energetic states and signs of increased cell membrane turnover in regions of visible tumor and differences to and between the “normal-appearing” brains of glioblastoma patients and the brains of healthy volunteers. Our pilot study confirmed the feasibility of the method, so differences between various genetic mutations or clinical applicability for follow-up monitoring can be assessed in larger cohorts.

Abstract

Background: Glioblastoma multiforme (GBM) is a highly malignant primary brain tumor with infiltration of, on conventional imaging, normal-appearing brain parenchyma. Phosphorus magnetic resonance spectroscopy (31P-MRS) enables the investigation of different energy and membrane metabolites. The aim of this study is to investigate regional differences of 31P-metabolites in GBM brains. Methods: In this study, we investigated 32 patients (13 female and 19 male; mean age 63 years) with naïve GBM using 31P-MRS and conventional MRI. Contrast-enhancing (CE), T2-hyperintense, adjacent and distant ipsilateral areas of the contralateral brain and the brains of age- and gender-matched healthy volunteers were assessed. Moreover, the 31P-MRS results were correlated with quantitative diffusion parameters. Results: Several metabolite ratios between the energy-dependent metabolites and/or the membrane metabolites differed significantly between the CE areas, the T2-hyperintense areas, the more distant areas, and even the brains of healthy volunteers. pH values and Mg²⁺ concentrations were highest in visible tumor areas and decreased with distance from them. These results are in accordance with the literature and correlated with quantitative diffusion parameters. Conclusions: This pilot study shows that 31P-MRS is feasible to show regional differences of energy and membrane metabolism in brains with naïve GBM, particularly between the different “normal-appearing” regions and between the contralateral hemisphere and healthy controls. Differences between various genetic mutations or clinical applicability for follow-up monitoring have to be assessed in a larger cohort.

Primary hypothalamic lymphoma with clinical findings mimicking pituitary apoplexy: a case report

BACKGROUND

Primary central nervous system lymphoma (PCNSL) is a rare but well-known extra-nodal lymphoma, which usually presents with non-Hodgkin B-cell lymphomas. PCNSL is generally located around the ventricle and is often detected as multiple lesions. It is rarely seen in the area of the hypothalamus.

CASE PRESENTATION

We report the case of a 48-year-old Caucasian woman with progressive short-term memory deterioration, headache, mental confusion, diabetes insipidus (DI) and hypopituitarism. Early findings were suggestive of a pituitary apoplexy. The results of tests performed during the initial admission at the tertiary health center revealed hypernatremia, hypopituitarism and DI. Intravenous hydrocortisone treatment was initiated for the secondary adrenal insufficiency, and 75 mcg/day of levothyroxine was started for the secondary hypothyroidism on the fourth day following hydrocortisone treatment. A daily dose of 120 mg desmopressin melt tablet was started twice a day for polyuria/polydipsia after the patient's volume status was balanced. A brain magnetic resonance imaging scan revealed a mass lesion in the hypothalamic area, which was surrounded by marked edema. Anti-edema treatment was initially started considering the suggestion by our neurosurgery team. The patient's clinical and laboratory findings improved after the initiation of the anti-edema therapy. Afterwards, a biopsy was performed, which diagnosed a malignant diffuse large B-cell lymphoma. Subsequently, intravenous high-dose methotrexate-based therapy was started; however, after the second cycle of chemotherapy, the patient died due to sepsis.

CONCLUSION

In this report, we present a case of hypopituitarism that developed due to the mass effect of hypothalamic lymphoma with clinical findings of pituitary apoplexy. Intracranial masses may cause obvious endocrinological findings related to hypopituitarism, while vague findings may also be observed due to partial failure. Therefore, it is important to perform a comprehensive endocrinological examination at the time of diagnosis in patients with intracranial masses.

Radiomic Based Machine Learning Performance for a Three Class Problem in Neuro-Oncology: Time to Test the Waters?

Simple Summary

Prior radiomic studies have addressed a two-class tumor classification problem (glioblastoma (GBM) versus primary CNS lymphoma (PCNSL) or GBM versus metastasis). However, this approach is prone to bias and excludes other common brain tumor types. We addressed a real-life clinical problem by including the three most common brain tumor types (GBM, PCNSL, and metastasis). We investigated two key issues using different MRI sequence combinations: performance variation based on tumor subregions (necrotic, enhancing, edema and combined enhancing, and necrotic masks), and performance metrics based on the chosen classifier model/feature selection combination. Our study provides evidence that radiomics-based three-class tumor differentiation is feasible, and that embedded models perform better than those with a priori feature selection. We found that T1 contrast enhanced is the single best sequence with comparable performance to that of multiparametric MRI, and model performance varies based on tumor subregion and the combination of model/feature selection methods.

Abstract

Prior radiomics studies have focused on two-class brain tumor classification, which limits generalizability. The performance of radiomics in differentiating the three most common malignant brain tumors (glioblastoma (GBM), primary central nervous system lymphoma (PCNSL), and metastatic disease) is assessed; factors affecting the model performance and usefulness of a single sequence versus multiparametric MRI (MP-MRI) remain largely unaddressed. This retrospective study included 253 patients (120 metastatic (lung and brain), 40 PCNSL, and 93 GBM). Radiomic features were extracted for whole a tumor mask (enhancing plus necrotic) and an edema mask (first pipeline), as well as for separate enhancing and necrotic and edema masks (second pipeline). Model performance was evaluated using MP-MRI, individual sequences, and the T1 contrast enhanced (T1-CE) sequence without the edema mask across 45 model/feature selection combinations. The second pipeline showed significantly high performance across all combinations (Brier score: 0.311–0.325). GBRM fit using the full feature set from the T1-CE sequence was the best model. The majority of the top models were built using a full feature set and inbuilt feature selection. No significant difference was seen between the top-performing models for MP-MRI (AUC 0.910) and T1-CE sequence with (AUC 0.908) and without edema masks (AUC 0.894). T1-CE is the single best sequence with comparable performance to that of multiparametric MRI (MP-MRI). Model performance varies based on tumor subregion and the combination of model/feature selection methods.

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.

Radiomics-Based Differentiation between Glioblastoma, CNS Lymphoma, and Brain Metastases: Comparing Performance across MRI Sequences and Machine Learning Models

Prior radiomics studies have focused on two-class brain tumor classification, which limits generalizability. The performance of radiomics in differentiating the three most common malignant brain tumors (glioblastoma (GBM), primary central nervous system lymphoma (PCNSL), and metastatic disease) is assessed; factors affecting the model performance and usefulness of a single sequence versus multiparametric MRI (MP-MRI) remain largely unaddressed. This retrospective study included 253 patients (120 metastatic (lung and brain), 40 PCNSL, and 93 GBM). Radiomic features were extracted for whole a tumor mask (enhancing plus necrotic) and an edema mask (first pipeline), as well as for separate enhancing and necrotic and edema masks (second pipeline). Model performance was evaluated using MP-MRI, individual sequences, and the T1 contrast enhanced (T1-CE) sequence without the edema mask across 45 model/feature selection combinations. The second pipeline showed significantly high performance across all combinations (Brier score: 0.311–0.325). GBRM fit using the full feature set from the T1-CE sequence was the best model. The majority of the top models were built using a full feature set and inbuilt feature selection. No significant difference was seen between the top-performing models for MP-MRI (AUC 0.910) and T1-CE sequence with (AUC 0.908) and without edema masks (AUC 0.894). T1-CE is the single best sequence with comparable performance to that of multiparametric MRI (MP-MRI). Model performance varies based on tumor subregion and the combination of model/feature selection methods.

From Neurosurgical Planning to Histopathological Brain Tumor Characterization: Potentialities of Arcuate Fasciculus Along-Tract Diffusion Tensor Imaging Tractography Measures

Background: Tractography has been widely adopted to improve brain gliomas' surgical planning and guide their resection. This study aimed to evaluate state-of-the-art of arcuate fasciculus (AF) tractography for surgical planning and explore the role of along-tract analyses in vivo for characterizing tumor histopathology.

Methods: High angular resolution diffusion imaging (HARDI) images were acquired for nine patients with tumors located in or near language areas (age: 41 ± 14 years, mean ± standard deviation; five males) and 32 healthy volunteers (age: 39 ± 16 years; 16 males). Phonemic fluency task fMRI was acquired preoperatively for patients. AF tractography was performed using constrained spherical deconvolution diffusivity modeling and probabilistic fiber tracking. Along-tract analyses were performed, dividing the AF into 15 segments along the length of the tract defined using the Laplacian operator. For each AF segment, diffusion tensor imaging (DTI) measures were compared with those obtained in healthy controls (HCs). The hemispheric laterality index (LI) was calculated from language task fMRI activations in the frontal, parietal, and temporal lobe parcellations. Tumors were grouped into low/high grade (LG/HG).

Results: Four tumors were LG gliomas (one dysembryoplastic neuroepithelial tumor and three glioma grade II) and five HG gliomas (two grade III and three grade IV). For LG tumors, gross total removal was achieved in all but one case, for HG in two patients. Tractography identified the AF trajectory in all cases. Four along-tract DTI measures potentially discriminated LG and HG tumor patients (false discovery rate < 0.1): the number of abnormal MD and RD segments, median AD, and MD measures. Both a higher number of abnormal AF segments and a higher AD and MD measures were associated with HG tumor patients. Moreover, correlations (unadjusted p < 0.05) were found between the parietal lobe LI and the DTI measures, which discriminated between LG and HG tumor patients. In particular, a more rightward parietal lobe activation (LI < 0) correlated with a higher number of abnormal MD segments (R = −0.732) and RD segments (R = −0.724).

Conclusions: AF tractography allows to detect the course of the tract, favoring the safer-as-possible tumor resection. Our preliminary study shows that along-tract DTI metrics can provide useful information for differentiating LG and HG tumors during pre-surgical tumor characterization.

Multi-parametric MR Imaging Biomarkers Associated to Clinical Outcomes in Gliomas: A Systematic Review

PURPOSE

To systematically review evidence regarding the association of multiparametric biomarkers with clinical outcomes and their capacity to explain relevant subcompartments of gliomas.

MATERIALS AND METHODS

Scopus database was searched for original journal papers from January 1st, 2007 to February 20th, 2017 according to PRISMA. Four hundred forty-nine abstracts of papers were reviewed and scored independently by two out of six authors. Based on those papers we analyzed associations between biomarkers, subcompartments within the tumor lesion, and clinical outcomes. From all the articles analyzed, the twenty-seven papers with the highest scores were highlighted to represent the evidence about MR imaging biomarkers associated with clinical outcomes. Similarly, eighteen studies defining subcompartments within the tumor region were also highlighted to represent the evidence of MR imaging biomarkers. Their reports were critically appraised according to the QUADAS-2 criteria.

RESULTS

It has been demonstrated that multi-parametric biomarkers are prepared for surrogating diagnosis, grading, segmentation, overall survival, progression-free survival, recurrence, molecular profiling and response to treatment in gliomas. Quantifications and radiomics features obtained from morphological exams (T1, T2, FLAIR, T1c), PWI (including DSC and DCE), diffusion (DWI, DTI) and chemical shift imaging (CSI) are the preferred MR biomarkers associated to clinical outcomes. Subcompartments relative to the peritumoral region, invasion, infiltration, proliferation, mass effect and pseudo flush, relapse compartments, gross tumor volumes, and highrisk regions have been defined to characterize the heterogeneity. For the majority of pairwise cooccurrences, we found no evidence to assert that observed co-occurrences were significantly different from their expected co-occurrences (Binomial test with False Discovery Rate correction, α=0.05). The co-occurrence among terms in the studied papers was found to be driven by their individual prevalence and trends in the literature.

CONCLUSION

Combinations of MR imaging biomarkers from morphological, PWI, DWI and CSI exams have demonstrated their capability to predict clinical outcomes in different management moments of gliomas. Whereas morphologic-derived compartments have been mostly studied during the last ten years, new multi-parametric MRI approaches have also been proposed to discover specific subcompartments of the tumors. MR biomarkers from those subcompartments show the local behavior within the heterogeneous tumor and may quantify the prognosis and response to treatment of gliomas.

Presurgical Identification of Primary Central Nervous System Lymphoma with Normalized Time-Intensity Curve: A Pilot Study of a New Method to Analyze DSC-PWI

BACKGROUND AND PURPOSE

DSC-PWI has demonstrated promising results in the presurgical diagnosis of brain tumors. While most studies analyze specific parameters derived from time-intensity curves, very few have directly analyzed the whole curves. The aims of this study were the following: 1) to design a new method of postprocessing time-intensity curves, which renders normalized curves, and 2) to test its feasibility and performance on the diagnosis of primary central nervous system lymphoma.

MATERIALS AND METHODS

Diagnostic MR imaging of patients with histologically confirmed primary central nervous system lymphoma were retrospectively reviewed. Correlative cases of glioblastoma, anaplastic astrocytoma, metastasis, and meningioma, matched by date and number, were retrieved for comparison. Time-intensity curves of enhancing tumor and normal-appearing white matter were obtained for each case. Enhancing tumor curves were normalized relative to normal-appearing white matter. We performed pair-wise comparisons for primary central nervous system lymphoma against the other tumor type. The best discriminatory time points of the curves were obtained through a stepwise selection. Logistic binary regression was applied to obtain prediction models. The generated algorithms were applied in a test subset.

RESULTS

A total of 233 patients were included in the study: 47 primary central nervous system lymphomas, 48 glioblastomas, 39 anaplastic astrocytomas, 49 metastases, and 50 meningiomas. The classifiers satisfactorily performed all bilateral comparisons in the test subset (primary central nervous system lymphoma versus glioblastoma, area under the curve = 0.96 and accuracy = 93%; versus anaplastic astrocytoma, 0.83 and 71%; versus metastases, 0.95 and 93%; versus meningioma, 0.93 and 96%).

CONCLUSIONS

The proposed method for DSC-PWI time-intensity curve normalization renders comparable curves beyond technical and patient variability. Normalized time-intensity curves performed satisfactorily for the presurgical identification of primary central nervous system lymphoma.

Conventional MR and DW imaging findings of cerebellar primary CNS lymphoma: comparison with high-grade glioma

Primary central nervous system lymphomas (PCNSLs) and high-grade gliomas (HGGs) arising in the cerebellum is extremely low, making the differential diagnosis difficult or even impossible. The purpose of this study was to define the MR features of cerebellar PCNSL in immunocompetent patients, and to determine whether a combination of conventional MR and DW imaging can assist in the differentiation of PCNSLs and HGGs. Twelve PCNSLs and 15 HGGs confirmed by pathological analysis were retrospectively identified. The apparent diffusion coefficient (ADC) and conventional MRI parameters were compared for differences between PCNSL and HGG groups using the independent sample t test or chi-square test. Both ADC_min and ADC_total values were lower in the PCNSL group than those in the HGG group (ADC_min: 0.53 × 10⁻³ vs. 0.83 × 10⁻³ mm²/sec, P < 0.001; ADC_total: 0.66 × 10⁻³ vs. 0.98 × 10⁻³ mm²/sec, P = 0.001). As for conventional MR features, there were significant difference in the tumor size, enhancement patterns, the presence of cystic changes, edema degree and streak-like edema (all P < 0.01); but there were no significant difference in lesion type, the presence of bleeding, and involvement of brain surface between two groups (P = 0.554, 0.657 and 0.157, respectively). The results revealed that several conventional MR features, including enhancement patterns, branch-like enhancement and streak-like edema may be useful for the differentiation of PCNSL and HGG in cerebellum and, when combined with ADC values, further improve the discriminating ability.

The role of MRI in the diagnosis and treatment of gastric cancer

Gastric cancer is the fifth most common malignancies and the third leading cause of cancer-related death worldwide, with more than 40% of new cases occurring in China. With the advancement of treatment methods, the application of adjuvant therapy and targeted drugs, the prognosis of patients with gastric cancer has been significantly improved. In recent years, more and more studies have reported that magnetic resonance imaging (MRI) showed great value in the clinical application among patients with gastric cancer, including preoperative staging, treatment response evaluation, predicting prognosis and histopathological features, treatment guidance, and molecular imaging. The remarkable research progress of MRI in gastric cancer will provide new evaluation and treatment approaches for clinical diagnosis and treatment. This article aims to review the current status of the application and research progress of MRI in patients with gastric cancer.

Differentiation among Glioblastomas, Primary Cerebral Lymphomas, and Solitary Brain Metastases Using Diffusion-Weighted Imaging and Diffusion Tensor Imaging: A PRISMA-Compliant Meta-analysis

Previous studies showed a high diagnostic value of diffusion-weighted imaging (DWI) and diffusion tensor imaging (DTI) in differentiation among glioblastomas, primary cerebral lymphomas (PCLs), and solitary brain metastases, whereas other studies reported a low or no diagnostic value of DWI and DTI in differentiation among the three types of brain malignant tumors. In order to enhance the strength of evidence, meta-analysis was conducted to summarize results of studies evaluating the diagnostic values of DWI or DTI in differentiation among the three types of brain malignant tumors. Articles evaluating the diagnostic values of DWI or DTI in differentiation among the three types of tumors and published before December 2019 were searched in databases (PubMed, Medline, Web of Science, EMBASE, and Google Scholar). A summary of sensitivity, specificity, positive likelihood ratios (PLR), negative likelihood ratios (NLR), and diagnostic odds ratio (DOR) were calculated from the true positive (TP), true negative (TN), false positive (FP), and false negative (FN) of each study using STATA 12.0 software and Meta-Disc Version 1.4. In addition, the summary receive-operating characteristic (SROC) curve was constructed. Ultimately, we included 19 diagnostic studies (including 735 glioblastomas patients, 31 PCLs patients, and 792 patients with solitary brain metastases). Regarding differentiation between glioblastomas and solitary brain metastases using DWI or DTI, the calculated pooled parameters were as follows: sensitivity, 0.84 [95% confidence interval (CI): 0.78-0.89]; specificity, 0.88 (95% CI: 0.83-0.92); PLR, 7.2 (95% CI: 4.6-11.3); NLR, 0.18 (95% CI: 0.12-0.27); and DOR, 41 (95% CI: 18-93). The analysis showed a significant heterogeneity (sensitivity, I² = 91.31%, p < 0.01; specificity, I² = 89.24%, p < 0.01). In conclusion, DWI and DTI showed a moderate diagnostic value for differentiating glioblastomas from solitary brain metastasis. Additionally, large-scale prospective studies are essential to explore differentiation between PCLs and solitary brain metastases using DWI or DTI.

Multi-Disease Segmentation of Gliomas and White Matter Hyperintensities in the BraTS Data Using a 3D Convolutional Neural Network

An important challenge in segmenting real-world biomedical imaging data is the presence of multiple disease processes within individual subjects. Most adults above age 60 exhibit a variable degree of small vessel ischemic disease, as well as chronic infarcts, which will manifest as white matter hyperintensities (WMH) on brain MRIs. Subjects diagnosed with gliomas will also typically exhibit some degree of abnormal T2 signal due to WMH, rather than just due to tumor. We sought to develop a fully automated algorithm to distinguish and quantify these distinct disease processes within individual subjects’ brain MRIs. To address this multi-disease problem, we trained a 3D U-Net to distinguish between abnormal signal arising from tumors vs. WMH in the 3D multi-parametric MRI (mpMRI, i.e., native T1-weighted, T1-post-contrast, T2, T2-FLAIR) scans of the International Brain Tumor Segmentation (BraTS) 2018 dataset (n_training = 285, n_validation = 66). Our trained neuroradiologist manually annotated WMH on the BraTS training subjects, finding that 69% of subjects had WMH. Our 3D U-Net model had a 4-channel 3D input patch (80 × 80 × 80) from mpMRI, four encoding and decoding layers, and an output of either four [background, active tumor (AT), necrotic core (NCR), peritumoral edematous/infiltrated tissue (ED)] or five classes (adding WMH as the fifth class). For both the four- and five-class output models, the median Dice for whole tumor (WT) extent (i.e., union of AT, ED, NCR) was 0.92 in both training and validation sets. Notably, the five-class model achieved significantly (p = 0.002) lower/better Hausdorff distances for WT extent in the training subjects. There was strong positive correlation between manually segmented and predicted volumes for WT (r = 0.96) and WMH (r = 0.89). Larger lesion volumes were positively correlated with higher/better Dice scores for WT (r = 0.33), WMH (r = 0.34), and across all lesions (r = 0.89) on a log(10) transformed scale. While the median Dice for WMH was 0.42 across training subjects with WMH, the median Dice was 0.62 for those with at least 5 cm³ of WMH. We anticipate the development of computational algorithms that are able to model multiple diseases within a single subject will be a critical step toward translating and integrating artificial intelligence systems into the heterogeneous real-world clinical workflow.