Keywords: Tumors (Pre-Treatment), Machine Learning/Artificial Intelligence

Motivation: A novel and noninvasive method for distinguishing intracranial solitary fibrous tumor (ISFT) from angiomatous meningioma (AM) and predicting patient outcomes is urgent.

Goal(s): To evaluate the value of a MRI-based deep learning radiomic nomogram (DLRN) in distinguishing ISFT from AM and predicting patient outcomes.

Approach: A MRI-based DLRN was developed on training cohort (TC). We then validated it's performance on external validation cohort (EVC). Moreover, we investigated the value of the DLRN in survival analysis.

Results: The performance of DLRN was excellent (0.86 [0.84–0.88]) on EVC. Besides, DLRN was significantly associated with the overall survival (OS) of patients (n=273).

Impact: The proposed DLRN can potentially provide a noninvasive method for neurosurgeon to offer decision support for developing personalized treatment plans and predicting patient outcomes.

Keywords: Tumors (Pre-Treatment), Metabolism, 13C NMR, stable isotopomer analysis, metabolic imaging

Motivation: Metabolism plays a key role in the growth and proliferation of brain tumors including aggressive meningiomas. We previously reported that meningiomas preferentially utilize acetate as a bioenergetic substrate. However, the metabolism of acetate in the presence of glucose is not well understood.

Goal(s): To investigate the simultaneous in vivo metabolism of acetate and glucose in meningiomas.

Approach: We infused [2-¹³C]acetate and [U-¹³C]glucose as metabolic tracers in meningioma patients to determine the relative utilization of both nutrients by meningiomas.

Results: Grade-II meningiomas utilize relatively less amount of glucose (grade-II: 0.6% vs. grade-I: ~5.4%) and more acetate than grade-I meningiomas (grade-II: 45.8% vs. grade-I: ~31.24%).

Impact: No chemotherapy is available for the treatment of meningiomas. The findings from this study will be helpful in designing targeted metabolic therapy for aggressive meningiomas using small molecule inhibitors (e.g., ACSS2 inhibitor) involved in acetate metabolism.

Keywords: Tumors (Pre-Treatment), Neuro

Motivation: DWI is of key importance in evaluating biological behavior of meningioma, but  image quality of conventional ssEPI-DWI is unsatisfactory due to susceptibility artifact near the skull.

Goal(s): Our goal was to compare the image quality of ssEPI-DWI and MUSE-DWI in meningiomas, and to compare diagnostic accuracy of them in meningiomas grading.

Approach: We used a 5-point Likert scale to assess image quality of DWI, and calculated SNR and CNR for quantitative evaluation. Combined models were constructed by using ADC histogram parameters extracted from whole tumor.

Results: MUSE-DWI significantly improved imaging quality of DWI, and showed a significantly higher diagnostic accuracy in meningioma grading.

Impact: Meningioma is the most commom intracranial tumours. This study revealed that MUSE-DWI, compared with ssEPI-DWI, can improve the imaging quality and grading accuracy of meningiomas, contributing to better clinical evaluation of meningiomas.

Keywords: Tumors (Pre-Treatment), Cancer, Sodium, Relaxometry

Motivation: Enhance the specificity of sodium (²³Na) imaging in brain tumors.

Goal(s): Distinguish intracellular and extracellular sodium distribution in glioblastoma.

Approach: We assessed the fraction of fast-relaxing sodium (F_T2*s) in 19 glioblastoma patients using a 3D radial ²³Na pulse sequence with six echoes at 7T.

Results: F_T2*s was high in normal appearing white matter resembling expected intracellular behavior, and low in necrotic regions, akin to fluid-like environments. Values for contrast-enhancing tumors and adjacent edema were in between. The differentiation between normal brain tissue and changes in glioblastoma underscores the potential of F_T2*s to improve the specificity of sodium imaging in brain tumors.

Impact: The study could facilitate the establishment of the fast-relaxing sodium fraction as a diagnostic tool in brain tumors, potentially improving specificity in sodium quantification.

Keywords: Tumors (Pre-Treatment), Brain

Motivation: Alterations in glutamatergic and GABAergic (gamma-aminobutyric acid) mechanisms render peritumoral neuronal networks of infiltrating glioma hyper-excitable and more prone to seizures. 

Goal(s): Glutamate, glutamine, glutathione (GSH), and GABA are therefore key metabolites in glioma-associated epilepsy.

Approach: Nowadays, J-editing MR spectroscopy is the primary technique in the detection of low-abundant metabolites (e.g., GSH, GABA) that overlap with more prominent signals. We used this technique combined with sLASER sequence (MEGA-sLASER) to improve localization accuracy and showed its reliability/repeatability for GSH and GABA quantification in glioblastoma, IDH-wildtype.

Results: This approach could foster our understanding of the biological effects of novel drugs targeting tumor-associated epilepsy.

Impact: MEGA-sLASER might improve the detectability and MRS localization accuracy of low-abundant metabolites (GSH, GABA) even in rather small, heterogeneous solid tumors. Here, we show the reliability/reproducibility of this method in the investigation of glioma-associated epilepsy in glioblastoma patients.

Keywords: Tumors (Pre-Treatment), CEST & MT

Motivation: Currently, APT and NOE imaging with extrapolated semi-solid magnetization transfer reference (EMR) fitting has been performed with single-slice acquisition. 

Goal(s): To demonstrate the feasibility of 3D APT and NOE imaging with EMR fitting using the 3D CEST sequence with CS-SENSE in patients with brain tumors.

Approach: 3D CEST imaging with CS-SENSE was adjusted to clinical scan. EMR fitting was performed on the 3D data. 

Results: 3D APT and NOE mapping was feasible in all patients with brain tumors by using the 3D CEST imaging with CS-SENSE within a clinically acceptable acquisition time.

Impact: The feasibility of 3D APT and NOE mapping using 3D CEST imaging with CS-SENSE was demonstrated. Quantitative evaluation of APT and NOE on multiple slices allows for the quantitative assessment for the entire tumor area.

Keywords: Tumors (Post-Treatment), Cancer, Deuterium metabolic Imaging

Motivation: Deuterium MRS imaging (DMRSI) can detect Warburg Effect in brain tumors; however, its clinical utility and value in brain tumor diagnosis and treatment has not been investigated.

Goal(s): To perform a preliminary investigation in human patients with brain tumor.

Approach: High-resolution dynamic DMRSI (HR-DMRSI) study was conducted in brain tumor patients on a 7T clinical scanner with an oral D66-glucose administration; biospecimens taken from DMRSI positive and negative regions were analyzed and compared with the DMRSI and clinical MRI results.

Results: HR-DMRSI technology enables detection and characterization of glioma infiltration with a level of precision surpassing traditional imaging modalities.

Impact: We clearly demonstrate that 7T high-resolution dynamic deuterium (²H) MRS imaging is able to detect and characterize glioma infiltration in individual human patients with high accuracy and specificity that exceeds conventional imaging modalities available in standard clinical setting.

Keywords: Tumors (Post-Treatment), DSC & DCE Perfusion, multiparametric brain tumor evaluation

Motivation: The lack of pathophysiologically relevant quantitative information hinders the precision management of brain tumors. Specifically, the characterization of intra-tumor heterogeneity influencing treatment decision-making, remains a critical challenge.

Goal(s): This work aims to optimize a recently developed technique MT-DICE, to provide multiparametric mapping information for more comprehensive brain tumor evaluation.

Approach: The MT-DICE technique was further refined by including 3D flow compensation, dictionary-based mapping, and water exchange quantification. Assessments for the repeatability of MT-DICE parameters and their agreement with routine measurements were performed.

Results: Excellent reproducibility and agreement were achieved. Spatially co-registered multiparametric maps from MT-DICE facilitated comprehensive brain tumor characterization.

Impact: Individuals with brain tumors may benefit from more comprehensive brain tumor evaluation using multiparametric maps derived from our improved MT-DICE technique.

Keywords: Tumors (Pre-Treatment), Radiotherapy, Glioblastoma

Motivation: GBM patients suffer from poor treatment outcome. The current radiotherapy planning (i.e. GTV, CTV, PTV) does not adjust the CTV margin to account for the microscopic spread of the glioma tumour cells along white matter fiber tracts.

Goal(s): Utilizing DTI MRI to tailor the CTV margin to each patient's unique tumour progression pathway.

Approach: Pre-surgical DTI-based tractography was used to quantify tumor spread probability, producing the Tumour Spread Index (TSI) map, which was used to generate a personalized CTV.

Results: This proof-of-concept study showed that using DTI-based tractography with a TSI map to personalize CTV improved coverage of recurrent regions in follow-up imaging.

Impact: Utilizing tractography to map the probable path of tumour spread and use this to direct radiation is a new paradigm in targeted radiotherapy which may lead to improved progression free survival in GBM patients.

Keywords: Tumors (Post-Treatment), Cancer, Glioblastoma, apparent diffusion coefficient, response assessment

Motivation: Early response assessment for glioblastoma may be possible using regions of low apparent diffusion coefficient (low-ADC) during MRI-linear accelerator (MR-Linac) treatment, but low-ADC definition has not been optimized.

Goal(s): Optimize definition of low-ADC for correlation with progression-free survival.

Approach: We defined low-ADC regions from near-daily diffusion-weighted imaging and weekly contrast-enhanced T₁-weighted imaging for 41 glioblastoma patients during MR-Linac treatment (3-6 weeks). We compared correlation strength across b-values (800 versus 2000s/mm²) and ADC thresholds (0.7 to 2.0μm²/ms).

Results: The optimal b-value/threshold combination was b=800s/mm² and 1.2μm²/ms (correlation for weeks 2-6).

Impact: We showed that early response assessment in glioblastoma during radiotherapy is possible with weekly acquisition of ADC maps and contrast-enhanced T₁-weighted imaging on MR-Linacs. Low-ADC regions could serve as targets for radiotherapy dose escalation to potentially extend patient survival.

Keywords: Tumors (Post-Treatment), Brain Connectivity, Neuroscience, Tumors

Motivation: Brain surgery glioma patients frequently exhibit attentional deficit, however the prediction of its appearance is largely unknown.

Goal(s): Investigate longitudinal temporal properties of executive function to gain insights about its relationship with cognitive attentive performance in gliomas.

Approach: We used longitudinal dynamic functional connectivity analysis of executive networking to associate it with neuropsychological attentive and executive performance.

Results: Post-surgical attentive performance is strictly related to functional temporal properties of executive networks, regardless of gliomas' features. Underlying substrates of impairment in the executive domain could be explained by looking at changes in temporal persistence of highly co-activated fronto-parietal networking.

Impact: Co-activation patterns framework enables the prediction of post-operative attentional deficits by looking at pre-surgical temporal features of executive networking. We demonstrate how the dynamic nature of the brain contains crucial features to develop clinically relevant imaging markers for gliomas recovery.

Keywords: Tumors (Post-Treatment), Tumor, glioma, neuro-oncology

Motivation: SOX2 is a marker of pluripotency that highlights tumor invasion beyond gross histological signatures, but the imaging characteristics of pluripotent tumor areas are unknown.

Goal(s): Do conventional imaging signatures deliniate SOX2 positive staining tumor cell regions?

Approach: This study compared imaging data from 22 glioma patients to aligned SOX2 autopsy tissue samples.

Results: SOX2-positive regions were often present beyond the imaging-defined tumor mass, with a positive but not collinear association between SOX2 and cell density across most. However, MR intensity distributions did not effectively distinguish SOX2 positivity.

Impact: These results highlight a novel signature of tumor presence that exists well-beyond the imaging-defined margin and is not readily detectable via conventional imaging. These areas are spared treatment and require new technological developments to detect non-invasively.

Keywords: Tumors (Pre-Treatment), Brain, Gliomas, Isocitrate Dehydrogenase, Synthetic MRI, Histogram Analysis

Motivation: There is an urgent need to identify a novel, cost-effective, and non-invasive method for determining the IDH mutation status in differentiating between astrocytoma and glioblastoma.

Goal(s): To investigate the potential value of whole-tumor histogram metrics derived from synthetic MRI in distinguishing IDH mutation status.

Approach: Histogram metrics were extracted from the quantitative maps. Variables with statistical significance in univariate analysis were included in multivariate logistic regression analysis to develop the combined model. The AUC were used to assess the diagnostic performance of metrics and models.

Results: The combined model could be a valuable preoperative tool to distinguish IDH mutation status.

Impact: The current study proposes a combined model that comprises T1-10th, cT1-10th, and age. This model demonstrates differentiation between IDH-M astrocytoma and IDH-W glioblastoma. Moreover, it has the potential to decrease genetic testing expenses while offering treatment decision support for clinicians.

Keywords: Tumors (Pre-Treatment), CEST & MT, Asymmetry-based and Lorentzian-fit-based CEST contrast reconstruction

Motivation: There is a scarcity of studies comparing the clinical value of asymmetry- and different Lorentzian-fit-based CEST contrasts of the amide proton transfer (APT) and semi-solid magnetization transfer (ssMT) in patients with glioma.

Goal(s): To assess and compare the potential of asymmetry-based (APTw_asym) and Lorentzian-fit-based CEST imaging with (MTR_RexAPT and MTR_RexMT) and without (MT_const) relaxation compensation for the prediction of therapy response and survival in patients with glioma.

Approach: 78 study participants prospectively underwent CEST MRI at baseline before radiotherapy.

Results: Imaging of the MTR_RexAPT and MTR_RexMT predicted response to radiotherapy, whilst the MTR_RexAPT was also associated with progression-free survival.

Impact: Here we demonstrate for the first time that Lorentzian-fit-based CEST imaging of the APT and ssMT with relaxation compensation can predict therapy response and progression-free survival of patients with glioma at baseline before radiotherapy, at 3T.

Keywords: Tumors (Post-Treatment), Hyperpolarized MR (Non-Gas)

Motivation: Despite aggressive treatments, patients with GBM have a median overall survival of 14-16 months and a need for noninvasive evaluation of therapeutics is apparent.  

Goal(s): To assess whether treatment-induced metabolic changes can be observed in patients using parameters derived from hyperpolarized 13C metabolic imaging data. 

Approach: 19 patients with recurrent GBM were followed for at least 6 months following treatment initiation and evaluated at various timepoints using HP-13C imaging. 

Results: A difference in trends following treatment was observed in pyruvate-to-lactate conversion for patients who received anti-angiogenic treatments as compared to those who received a protein kinase inhibitor.

Impact: Given the challenges associated with evaluating progression and response to therapy in patients with glioblastoma using conventional MRI, this study provided evidence that hyperpolarized carbon-13 techniques can detect serial changes in dynamic metabolism which might help predict disease status. 

Keywords: Tumors (Post-Treatment), Spectroscopy, Tumor, Drugs, Hyperpolarized MR (Non-Gas)

Motivation: Inhibiting TERT or its upstream transcription factor GABPB1 can result in tumor growth inhibition. Inhibiting EGFR, upstream of TERT, can also reduce TERT expression.

Goal(s): Investigating the combined effects of EGFR and TERT inhibition and assessing whether our MRS-based biomarkers can detect the impact of this combination therapy in cell and animal models.

Approach: Proton and hyperpolarized ¹³C spectroscopy in cell and animal models

Results: Enhanced inhibition of both cell and tumor growth was observed in our GBM models when TERT/GABPB1 and EGFR were targeted simultaneously. This was associated with a drop in hyperpolarized lactate production from pyruvate.

Impact: This study identifies HP lactate as a metabolic biomarker of response to the dual TERT/GABPB1 and EGFR inhibition in cells and animals and points to the value of this biomarker in detecting the added value of this novel combination therapy.

Keywords: Tumors (Pre-Treatment), Brain Connectivity, Glioma; Multimodal; Integration; Structural Connectivity; Functional Connectivity

Motivation: Brain networks glioma’s disruption was often explored through separate examinations of structural and functional connectivity. However, there were limited efforts in glioma research to investigate the interplay between structure-function and how this connection might influence our comprehension.

Goal(s): Can integrating structural and functional connectivity aid understanding the alterations’ neurobiological foundation in brain networks caused by glioma?

Approach: The study design involves standard diffusion MR and rs-fMRI preprocessing, statistical methods including Pearson and Spearman correlation and Euclidean distance computation.

Results: This study underscores the significance of examining structure-function integration, where both microstructure and function play crucial roles in relation to white matter integrity. 

Impact: Glioma, the primary brain tumor, affects both structural and functional connectivity. Understanding alterations in structure-function integration and connection with single-modalities, may be of highest significance for a more comprehensive explanation of compensatory mechanisms induced by glioma and its clinical progression.

Keywords: Tumors (Pre-Treatment), Tumor, Glioma; SWI; ADC; Machine learning

Motivation: WHO Grade, IDH mutation and MGMT promoter methylation are important for precise diagnosis and treatment plans for diffuse glioma patients. 

Goal(s): This study aimed to investigate the predictive value of radiomics features extracted from Structural MRI, ADC and SWI. 

Approach: Radiomic features were extracted from T1WI, T2WI, T1CE, FLAIR, ADC and SWI. Analysis of variance F-test were used for feature selection. 11 classifiers were utilized for model establishment.

Results: For WHO Grade task, the highest AUC was 0.990;   for IDH mutation task, the highest AUC was 0.947. All the constructed models failed to predict MGMT promoter methylation status efficiently. 

Impact: This work will help neuro-oncologists better understand the radiological manifestation of gliomas.

Keywords: Tumors (Post-Treatment), Arterial spin labelling, Machine Learning/Artificial Intelligence, Tumor

Motivation: The accuracy of cerebral blood flow (CBF) quantification in arterial spin labelling (ASL) may reduce in regions exhibiting pathology-induced signal abnormalities in proton density (PD) images.

Goal(s): To develop an algorithm for improved CBF quantification by correcting signal abnormalities in PD images due to pathology.

Approach: To correct signal abnormalities, an image-inpainting algorithm based on deep learning (DL) was developed using healthy subject data. The algorithm was demonstrated with an application to patients post tumour treatment.

Results: The developed DL algorithm was able to effectively correct signal abnormalities, resulting in improved CBF maps.

Impact: The improvement in CBF accuracy through DL-corrected PD images may aid clinicians in their assessment of patients. This study demonstrates the potential benefit of the proposed method in an example application of monitoring tumour recurrence post treatment with ASL.

Keywords: Tumors (Pre-Treatment), Radiomics, Malignant gliomas, Habitat imaging

Motivation: Malignant gliomas are characterized by considerable intra-tumor heterogeneity directly related to treatment failure. Habitat imaging allows to visualize in vivo such heterogeneity.

Goal(s): Investigating a novel approach based on 3T PET/MRI acquisitions for assessing HYpoxia, PErfusion, DIffusion, and methionine-PET for tissue metabolism in malignant gliomas. 

Approach: Quantitative data from PET/MRI are combined in a unique HYPERDIrect map to identify discrete radiomic clusters or `habitats', possibly reflecting diverse genetic and biological features, to be proven by image-guided tissue sampling.

Results: Preliminary analysis showed high habitat imaging reproducibility and a reliable correlation between the expected microenvironment of the different habitats and the actual histopathological characteristics.

Impact: The HYPERDIrect map may represent a tool for decoding pattern of glioma heterogeneity and a new biomarker for stratifying prognosis and selecting patients for personalized approaches.