Abstract Background Urinary tract obstruction (UTO) can lead to progressive renal impairment, making accurate evaluation of split renal function (SRF) essential for clinical decision-making. Although radionuclide renal dynamic scintigraphy remains the gold standard for SRF assessment, its clinical application is constrained by procedural complexity, limited availability, and sensitivity to anatomical variations. Thus, there is a clinical need for simpler, reliable, and noninvasive alternative approaches. This study aimed to develop and validate a predictive model for SRF grading by integrating clinical variables, laboratory parameters, and quantitative contrast-enhanced computed tomography (CECT) features in patients with UTO. Methods A retrospective cohort of 78 patients with UTO (150 kidneys) was analyzed. Based on split renal glomerular filtration rate (GFR) determined using the Gates method, kidneys were categorized into normal, mild-to-moderate impairment, and severe impairment groups. Clinical variables, laboratory parameters, and quantitative CECT features were collected. Univariate and multivariate logistic regression analyses were performed to identify independent predictors and construct SRF grading models. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and decision curve analysis (DCA). Results In Task 1 (normal vs. abnormal) and Task 2 (normal vs. mild-to-moderate), age was the key univariate predictor, while hemoglobin (Hb) and renal cortical thickness (Rc) were identified as independent predictors in the laboratory-based and CECT-based multivariate models, respectively. For Task 1, the Combined model [Clinical predictor (Age) + Laboratory model (Hb-based) + CECT model (Rc-based)] achieved the highest diagnostic performance (AUC = 0.890); for Task 2, the optimal model was [Clinical predictor (Age) + CECT model (Rc-based)] (AUC = 0.810). For Task 3 (mild-to-moderate vs. severe), Hb was the strongest univariate predictor, and Rc was the sole independent predictor in the CECT-based multivariate model. The highest diagnostic accuracy for Task 3 was achieved by the combined Laboratory predictor (Hb) + CECT model (Rc-based), with an AUC of 0.970. Conclusion The CECT model (Rc-based) serves as a crucial imaging biomarker for evaluating SRF impairment in patients with UTO. Task-specific models combining the CECT model (Rc-based) with a clinical predictor (Age) and laboratory information—either as a univariable predictor (Hb) or as a multivariable laboratory model (Hb-based)—showed superior predictive performance. This integrated, noninvasive strategy may serve as a useful adjunct to radionuclide imaging for individualized SRF assessment, pending prospective validation.

Metanephric adenoma (MA) and papillary renal cell carcinoma (pRCC) demonstrate overlapping hypovascular appearances on contrast-enhanced computed tomography (CECT), yet reliable quantitative imaging criteria for their preoperative differentiation are still lacking. This study aimed to evaluate the diagnostic value of quantitative parameters derived from multiphasic CECT for differentiating MA from pRCC. Ninety-three patients with pathologically confirmed renal tumors (MA, n = 31; pRCC, n = 62) were retrospectively included. CT attenuation values of renal tumors and the ipsilateral renal cortex were measured on unenhanced (UP), corticomedullary (CMP), and nephrographic (NP) phases in all 93 patients, and on the excretory (EP) phase in 84 patients (26 MA, 58 pRCC) as EP images were unavailable in the remaining nine patients. Multiple quantitative parameters were derived. After standardization, univariable and multivariable logistic regression analyses were performed to identify independent predictors, and several multivariable diagnostic models were constructed and compared. Univariable analysis demonstrated significant differences between pRCC and MA in age, sex, and multiple quantitative parameters derived from the NP and EP. Among these parameters, the absolute tumor enhancement difference between the excretory and unenhanced phases (EPT–UPT) showed the highest diagnostic performance [area under the curve (AUC) = 0.885, 95

To investigate the value of qualitative and quantitative contrast-enhanced CT (CECT) features for noninvasive identification of two distinct vascular patterns, vessels that encapsulate tumor clusters (VETC) and/or microvascular invasion (MVI), in solitary early-stage (BCLC 0-A) hepatocellular carcinoma (HCC) and assess their prognostic implications. We retrospectively included 347 patients with solitary early-stage HCC who underwent preoperative CECT and subsequent resection at two centers. Patients were divided into V/M+ (MVI and/or VETC positive, n = 174) and VM− (both MVI and VETC negative, n = 173) groups based on histopathology. Four predictive models (clinical, CT quantitative, CT qualitative, and combined) integrating clinical and CECT features were developed and validated for identifying V/M+ status. The optimal model was further applied to predict 2-year recurrence-free survival (RFS). Sensitivity analysis was performed using propensity score matching (PSM). Models’ performance was evaluated and compared using AUC analyses and DeLong tests. The combined model [serum AFP ≥ 200 ng/mL, non-smooth tumor margin, internal arteries, and lower tumor-to-liver density ratio in the portal venous phase (P-TLR)] achieved optimal predictive performance for V/M + HCC, with training AUC of 0.784 and 0.782 pre- and post-PSM, and external validating AUC of 0.794. A derived V/M+ score stratified patients, with higher scores associated with significantly shorter 2-year RFS. V/M+ score ≥ 34 and tumor size ≥ 60 mm were significant predictors of HCC recurrence (p < 0.05). The combined model integrating clinical and CECT-based features, enables non-invasive assessment of V/M status in early-stage solitary HCC and effectively stratifies patients according to recurrence risk. Specific CT-based qualitative and quantitative features are associated with a distinct vascular pattern of BCLC stage 0-A HCC. The developed combined model and derived V/M+ score offer a reliable tool for clinicians to predict V/M + HCC and patients’ 2-year RFS.

Magnetic resonance imaging (MRI) radiomics has shown promise in glioma grading and isocitrate dehydrogenase (IDH) mutation prediction, but traditional whole-tumor approaches overlook intratumoral heterogeneity, limiting diagnostic accuracy and interpretability. This study aims to explore cellularity habitat-based MRI radiomics for precise grading and IDH mutation status prediction in adult-type diffuse glioma (ADG). A total of 625 ADG patients were retrospectively collected. Whole-tumor volumes of interest (VOIs) were delineated on four conventional MRI sequences (T1WI, T2WI, T2-FLAIR, and CE-T1WI) and segmented into three cellularity habitats using apparent diffusion coefficient (ADC)-based K-means clustering: H1 (low ADC), H2 (medium ADC), and H3 (high ADC). Radiomic features were extracted from individual and combined habitats, and predictive models were developed using a disentangled-learning-based multi-sequence fusion network (DMSFN). Performance was evaluated using the area under the receiver operating characteristic curve (AUC), accuracy (ACC), sensitivity (SEN), and specificity (SPE). The optimal habitats for ADG grading (Grade 2 vs. Grade 3 + 4, Grade 2 + 3 vs. Grade 4) and IDH prediction were H1 + 2, H1 + 2, and H2 + 3, respectively. Combining T1WI, CE-T1WI, and T2-FLAIR sequences yielded the highest AUCs of 0.9360, 0.9605, and 0.8721 in the training set, and 0.8070, 0.8236, and 0.8180 in the independent test set. Shapley Additive exPlanation (SHAP) analysis identified key radiomic features contributing to model predictions, with CE-T1WI features consistently demonstrating high discriminative power. Integrating ADC-derived cellularity habitats with MRI radiomics significantly improves the accuracy and biological interpretability of ADG grading and IDH mutation status prediction, offering a robust, non-invasive approach for glioma characterization. Retrospectively registered.

Accurate classification of renal masses before treatment is crucial for therapeutic decision-making and patient outcome. This study developed and validated Multi-Phase Attention Network (MPANet), a multimodal deep learning model integrating multiphase contrast-enhanced CT and clinical information, which can utilize both complete-phase and missing-phase CT data for multiclass classification of four common and easily confusable renal tumors-clear cell renal cell carcinoma (ccRCC), papillary renal cell carcinoma (pRCC), oncocytic neoplasms (including chromophobe renal cell carcinoma (chRCC) and renal oncocytoma (RO)), and fat-poor angiomyolipoma (fpAML). A total of 1688 multi-center cases were enrolled. Across all test sets, MPANet consistently outperformed single-phase models. In the internal test set, MPANet achieved a macro-average AUC of 0.850, a micro-average AUC of 0.865, and an accuracy of 73.3%. These results compared favorably to assessments by four radiologists based on CT (accuracies 43.6-62.4%) and two radiologists using MRI with clear cell likelihood score (ccLS) system (accuracies 52.5% and 49.5%). The net improvement rate of MPANet over radiologist assessment ranged from 10.9% to 29.7%. In the two external test sets, macro-average AUCs were 0.811 and 0.813, and micro-average AUCs were 0.867 and 0.909, respectively. MPANet shows potential as a clinical decision-support tool for personalized renal tumor diagnosis.

To explore the predictive value of preoperative clinical and multiphase dynamic contrast-enhanced CT (CECT)-based qualitative and quantitative features for identifying Cytokeratin 19 (CK19)- and Glypican 3 (GPC3)-positive dual-phenotype hepatocellular carcinoma (DPHCC_CG+). A total of 363 HCC patients who received preoperative CECT and surgical resection from two medical centers were retrospectively included. Patients were divided into DPHCC_CG+ (CK19- and GPC3-positive) or Non-DPHCC_CG+ (CK19- and/or GPC3-negative) groups based on histopathology. Qualitative and quantitative CECT features, along with clinical variables, were compared between the two groups. Univariate and multivariate logistic regression analyses were performed to identify independent predictors of DPHCC_CG+. Three predictive models (CT qualitative-clinical, CT quantitative-clinical and combined model) were developed in the training cohort and externally validated. Model performance was assessed and compared using receiver operating characteristic (ROC) analysis and the DeLong test. Both the CT qualitative-clinical and CT quantitative-clinical models demonstrated good predictive performance for identifying DPHCC_CG + in both the training and external validation cohorts (AUCs > 0.79). Independent predictors of DPHCC_CG+ included serum AFP ≥ 400 ng/mL, presence of mosaic appearance, arterial phase tumor-to-aorta attenuation ratio (A-TAR ≤ 0.262), and delayed phase normalized washout ratio (D-NWR ≤ -0.141). The combined model integrating these predictors achieved optimal diagnostic performance, with AUCs of 0.878 (95

Incomplete MRI sequences pose a significant challenge to the reliability of multiparametric MRI (mp-MRI) radiomics models. This study aimed to develop a robust and noninvasive approach for accurate glioma grading and isocitrate dehydrogenase (IDH) mutation status prediction using incomplete MRI data. Conventional MRI scans of 2170 glioma patients were retrospectively collected from five clinical institutions and a public dataset. Radiomic features extracted from each sequence, including incomplete ones, were processed using a robust incomplete sequence estimation network (RISEN). This model imputes missing features and learns latent fusion representations for glioma grading and IDH mutation status prediction. Model performance was evaluated by determining the optimal combination of mp-MRI sequences and simulating various clinical scenarios with different rates of missing data, using the area under the curve (AUC). The optimal sequence combination of T1WI, CE-T1WI, and T2-FLAIR achieved the highest performance for glioma grading (AUC: 0.8160, 0.9136, 0.8031) and IDH prediction (AUC: 0.8657, 0.8731, 0.7682) in internal and external validation datasets with complete data. In simulated incomplete-sequence scenarios, RISEN exhibited only moderate declines for glioma grading (mean AUC: 0.7942, 0.8955, 0.7759) and IDH prediction (mean AUC: 0.8454, 0.8478, 0.7414). Comparable robustness was observed in real-world missing-sequence data (AUC: 0.8910 and 0.8854, respectively). RISEN demonstrates robust and clinically applicable performance for glioma grading and IDH mutation prediction across multiple cohorts, even under commonly encountered incomplete mp-MRI conditions. Question Incomplete MRI sequences present a major challenge for radiomics-based glioma grading and IDH mutation status prediction in real-world clinical settings. Findings The proposed RISEN model imputes missing features using latent representations, improving model robustness and maintaining high predictive performance across various missing data scenarios. Clinical relevance RISEN offers a reliable solution for real-world glioma diagnosis when MRI sequences are incomplete, ensuring consistent clinical decision-making and personalized therapy planning across diverse imaging conditions.

Blocking the C-X-C motif chemokine ligand-12/C-X-C motif chemokine receptor-4 (CXCL12/CXCR4) signal offers the potential to induce immunogenic cell death (ICD) and enhance immunotherapy of glioblastoma (GBM). However, traditional intracellular targeted delivery strategies and adenosine-mediated tumor immunosuppression limit its therapeutic efficacy. Herein, we present an acidity-triggered self-assembly nanoplatform based on bioorthogonal reaction to potentiate GBM immunotherapy through dual regulation of metabolism and immune pathways. AMD3100 (CXCR4 antagonist) and CPI-444 (adenosine 2A receptor inhibitor) were formulated into micelles, denoted as AMD@iNPDBCO and CPI@iNPN3, respectively. Upon administration, the pH-sensitive poly(2azepane ethyl methacrylate) group of AMD@iNPDBCO responds to the acidic tumor microenvironment, exposing the DBCO moiety, resulting in highly efficient bioorthogonal reaction with azide group on CPI@iNPN3 to form large-sized aggregates, ensuring extracellular drug release. The combination of AMD3100 and CPI-444 contributes to ICD induction, dendritic cell maturation, and immunosuppressive milieu alleviation by reducing tumor-associated macrophages, myeloid-derived suppressor cells, and regulatory T cells, leading to a robust antitumor response, thereby significantly prolonging survival in orthotopic GBM-bearing mice. Furthermore, the nanoplatform remarkably amplifies immuno-radiotherapy by potently evoking cytotoxic CD8+ T cell priming, and synergized with immune checkpoint blockade by delaying CD8+ T cell exhaustion. Our work highlights the potential of the in situ assembly nanoplatform tailored for delivery of extracellular-targeted therapeutic agents for boosting GBM immunotherapy.

The efficacy of checkpoint blockade immunotherapy for glioblastoma (GBM) is significantly influenced by the precise delivery of therapeutic agents that can penetrate the blood-brain barrier (BBB) and reprogram the tumor immune microenvironment. Conventional nanoscale carriers used for delivering immune checkpoint blockers are more likely to be internalized by tumor cells, leading to a loss of drug efficacy. This study presents a phosphatidylcholine (PC)-coated nanoparticle (PCNP) with an optimized PC ratio on its surface, achieving a balanced surface charge. This surface optimization minimizes nanoparticle-cell membrane interactions, reducing cellular uptake and thereby enhancing extracellular drug targeting efficacy. The constructed PC shell enabled PCNPs to penetrate the BBB mediated by choline transporters. The PC shell can attenuate interactions between PCNPs and cells, thereby preventing the internalization of PCNPs. Additionally, the poly-l-histidine core can undergo protonation in the acidic microenvironment, resulting in rapid disintegration of PCNPs and facilitating the quick release of the encapsulated CPI-444 (an extracellular adenosine receptor blocker) and temozolomide, inducing immunogenic cell death and blocking extracellular adenosine receptors to reverse the immunosuppressive feedback signaling pathway of the adenosinergic axis. This combination therapy has shown a novel therapeutic strategy for extracellular immune checkpoint blockade in GBM.

The integration of multimodal data, particularly medical images and tabular data encompassing physician-assessed radiological factors, holds significant promise for enhancing clinical decision-making. However, effective fusion of these heterogeneous data modalities remains challenging due to their disparate feature spaces and the limitations of current independent encoding approaches. We introduce FM-Bridge, a novel methodology leveraging vision-language foundation model (VLM) to address this challenge. Our approach capitalizes on the intrinsic imagetext embedding space alignment within VLMs to achieve robust multimodal fusion. We propose transforming clinical expertise-rich tabular data into semantically coherent textual descriptions, subsequently utilizing the VLM's text encoder to generate textual features explicitly aligned with image features. This method facilitates a more semantically congruent and effective fusion of medical image and tabular data, demonstrating potential for improved performance in downstream medical image analysis tasks compared to conventional methods. Code is available at https:// github.com/HKU- MedAI/FM-Bridge.

Clear cell renal cell carcinoma (ccRCC) exhibits significant biological heterogeneity, with aggressive forms demonstrating poor prognosis. Accurate preoperative discrimination between aggressive and indolent ccRCC is critical for individualized treatment but remains challenging. This study aimed to evaluate the performance of machine learning models based on multiparametric MRI radiomics for distinguishing aggressive from indolent ccRCC. This retrospective study included 157 patients with pathologically confirmed ccRCC, comprising 114 indolent and 43 aggressive cases. Regions of interest (ROIs) were manually delineated on five MRI sequences: T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), as well as the corticomedullary, nephrographic, and excretory phases of contrast-enhanced fat-suppressed T1WI (CE-fsT1WI). Thirty-one feature combinations derived from the five sequences were input into 168 classification models (constructed using 8 classifiers and 21 feature selection methods). The performance of 5,208 models was compared, and the top-ranked features were analyzed. Aggressive ccRCC showed significantly larger maximum tumor diameter compared with indolent tumors (8.3 [5.7–9.5] cm vs. 3.0 [2.2–4.2] cm, p < 0.05). Radiomic features derived from T2WI contributed most substantially to model performance relative to other MRI sequences, with the optimal classification model “RF + ICAP” achieving an area under the curve (AUC) of 0.960, accuracy of 86.1

The macrotrabecular-massive (MTM) subtype of hepatocellular carcinoma (HCC) is a histological variant with higher malignant potential. Non-invasive preoperative identification of MTM-HCC is crucial for precise treatment. Current radiomics-based diagnostic models often integrate multi-phase features by simple feature concatenation, which may inadequately explore the latent complementary information between phases. This study proposes a feature fusion-based radiomics model using multi-phase contrast-enhanced computed tomography (mpCECT) images. Features were extracted from the arterial phase (AP), portal venous phase (PVP), and delayed phase (DP) CT images of 121 HCC patients. The fusion model was constructed and compared against the traditional concatenation model. Five-fold cross-validation demonstrated that the feature fusion model combining AP and PVP features achieved the best classification performance, with an area under the receiver operating characteristic curve (AUC) of 0.839. Furthermore, for any combination of two phases, the feature fusion model consistently outperformed the traditional feature concatenation approach. In conclusion, the proposed feature fusion model effectively enhances the discrimination capability compared to traditional models, providing a new tool for clinical practice.

Pancreatic ductal adenocarcinoma (PDAC) with cystic features presents significant challenges in achieving an accurate preoperative diagnosis and in implementing appropriate clinical management. The aim of this study was to analyze the dynamic contrast-enhanced computed tomography (DCE-CT) findings of PDACs with cystic lesions and correlate them with histopathological findings. We retrospectively reviewed 40 patients with pathology-proven PDACs exhibiting cystic lesions who underwent preoperative DCE-CT imaging. The CT manifestations were classified into three subtypes based on the morphological characteristics of the cystic lesions: Type 1, small proportion (< 50

Data harmonization is critical for establishing generalizable model on multicenter medical data. Traditional data harmonization strategies aim to align data distributions from different sources, but often lack mechanisms to learn hidden complementary and discriminative manifestations from multicenter data. To this end, we proposed a methodology for harmonizing multicenter data by matching their first and second order statistics in a shared space, which is framed in an optimization architecture to learn harmonized and discriminative latent features for downstream classification modeling. The developed method integrated representation learning, feature dimension reduction and selection within a unified framework. Several relational regularizations such as data attribute preservation and feature-task correlation have been explored and incorporated to encourage learning potential associations inherent in multicenter data. Extensive evaluations on three independent clinical datasets have demonstrated the efficacy of the proposed method in producing harmonized and distinguishing data for multicenter medical prediction modeling.

The absence of MRI sequences is a common occurrence in clinical practice, posing a significant challenge for prediction modeling of non-invasive diagnosis of glioma (GM) via fusion of multi-sequence MRI. To address this issue, we propose a novel unified reciprocal assistance imputation-representation learning framework (namely REPAIR) for GM diagnosis modeling with incomplete MRI sequences. REPAIR facilitates a cooperative process between missing value imputation and multi-sequence MRI fusion by leveraging existing samples to inform the imputation of missing values. This, in turn, facilitates the learning of a shared latent representation, which reciprocally guides more accurate imputation of missing values. To tailor the learned representation for downstream tasks, a novel ambiguity-aware intercorrelation regularization is introduced to equip REPAIR by correlating imputation ambiguity and its impacts conveying to the learned representation via a fuzzy paradigm. Additionally, a multimodal structural calibration constraint is devised to correct for the structural shift caused by missing data, ensuring structural consistency between the learned representations and the actual data. The proposed methodology is extensively validated on eight GM datasets with incomplete MRI sequences and six clinical datasets from other diseases with incomplete imaging modalities. Comprehensive comparisons with state-of-the-art methods have demonstrated the competitiveness of our approach for GM diagnosis with incomplete MRI sequences, as well as its potential for generalization to various diseases with missing imaging modalities.

Quantifying individual deviations in brain morphology from normative references is useful for understanding neurodiversity and facilitating personalized management of brain health. Here we report Chinese brain normative references using morphological imaging scans of 24,061 healthy volunteers from 105 sites, revealing later peak ages of lifespan neurodevelopmental milestones (1.2-8.9 years) than European/North American populations. We model individual brain deviation scores in 3,932 individuals with different neurological disorders from population references to evaluate three key aspects of brain health assessment using machine learning approaches: estimating disease propensity, predicting cognitive and physical outcomes and assessing treatment effects with distinct disability progression. The norm-deviation scores outperformed raw structural measures in these evaluations. Chinese-specific normative brain references may foster personalized diagnosis and prognosis in neurological diseases, enabling clinically applicable assessments of brain health.

The co-release of Len and BMS-202 from iRGD-modified liposomes induced a synergistic antitumor immunotherapy, with early therapeutic monitoring by IVIM-MRI.

Activating the stimulator of the interferon gene (STING) is a promising immunotherapeutic strategy for converting "cold" tumor microenvironment into "hot" one to achieve better immunotherapy for malignant tumors. Herein, a manganese-based nanotransformer is presented, consisting of manganese carbonyl and cyanine dye, for MRI/NIR-II dual-modality imaging-guided multifunctional carbon monoxide (CO) gas treatment and photothermal therapy, along with triggering cGAS-STING immune pathway against triple-negative breast cancer. This nanosystem is able to transfer its amorphous morphology into a crystallographic-like formation in response to the tumor microenvironment, achieved by breaking metal-carbon bonds and forming coordination bonds, which enhances the sensitivity of magnetic resonance imaging. Moreover, the generated CO and photothermal effect under irradiation of this nanotransformer induce immunogenic death of tumor cells and release damage-associated molecular patterns. Simultaneously, the Mn acts as an immunoactivator, potentially stimulating the cGAS-STING pathway to augment adaptive immunity, resulting in promoting the secretion of type I interferon, the proliferation of cytotoxic T lymphocytes and M2-macrophages repolarization. This nanosystem-based gas-photothermal treatment and immunoactivating therapy synergistic effect exhibit excellent antitumor efficacy both in vitro and in vivo, reducing the risk of triple-negative breast cancer recurrence and metastasis; thus, this strategy presents great potential as multifunctional immunotherapeutic agents for cancer treatment.

To explore a subregion-based RadioFusionOmics (RFO) model for discrimination between adult-type grade 4 astrocytoma and glioblastoma according to the 2021 WHO CNS5 classification. 329 patients (40 grade 4 astrocytomas and 289 glioblastomas) with histologic diagnosis was retrospectively collected from our local institution and The Cancer Imaging Archive (TCIA). The volumes of interests (VOIs) were obtained from four multiparametric MRI sequences (T1WI, T1WI + C, T2WI, T2-FLAIR) using (1) manual segmentation of the non-enhanced tumor (nET), enhanced tumor (ET), and peritumoral edema (pTE), and (2) K-means clustering of four habitats (H1: high T1WI + C, high T2-FLAIR; (2) H2: high T1WI + C, low T2-FLAIR; (3) H3: low T1WI + C, high T2-FLAIR; and (4) H4: low T1WI + C, low T2-FLAIR). The optimal VOI and best MRI sequence combination were determined. The performance of the RFO model was evaluated using the area under the precision-recall curve (AUPRC) and the best signatures were identified. The two best VOIs were manual VOI3 (putative peritumoral edema) and clustering H34 (low T1WI + C, high T2-FLAIR (H3) combined with low T1WI + C and low T2-FLAIR (H4)). Features fused from four MRI sequences ( F_seq^1,2,3,4 ) outperformed those from either a single sequence or other sequence combinations. The RFO model that was trained using fused features F_seq^1,2,3,4 achieved the AUPRC of 0.972 (VOI3) and 0.976 (H34) in the primary cohort (p = 0.905), and 0.971 (VOI3) and 0.974 (H34) in the testing cohort (p = 0.402). The performance of subregions defined by clustering was comparable to that of subregions that were manually defined. Fusion of features from the edematous subregions of multiple MRI sequences by the RFO model resulted in differentiation between grade 4 astrocytoma and glioblastoma.