Magnetic Resonance Imaging of Primary Adult Brain Tumors: State of the Art and Future Perspectives

Nanophotonic-enhanced photoacoustic imaging for brain tumor detection

Photoacoustic brain imaging (PABI) has emerged as a promising biomedical imaging modality, combining high contrast of optical imaging with deep tissue penetration of ultrasound imaging. This review explores the application of photoacoustic imaging in brain tumor imaging, highlighting the synergy between nanomaterials and state of the art optical techniques to achieve high-resolution imaging of deeper brain tissues. PABI leverages the photoacoustic effect, where absorbed light energy causes thermoelastic expansion, generating ultrasound waves that are detected and converted into images. This technique enables precise diagnosis, therapy monitoring, and enhanced clinical screening, specifically in the management of complex diseases such as breast cancer, lymphatic disorder, and neurological conditions. Despite integration of photoacoustic agents and ultrasound radiation, providing a comprehensive overview of current methodologies, major obstacles in brain tumor treatment, and future directions for improving diagnostic and therapeutic outcomes. The review underscores the significance of PABI as a robust research tool and medical method, with the potential to revolutionize brain disease diagnosis and treatment.

Multimodal Image Confidence: A Novel Method for Tumor and Organ Boundary Representation

The indistinct boundaries of tumors and organs at risk in medical images present significant challenges in treatment planning and other tasks in radiation therapy. This study introduces an innovative analytical algorithm called multimodal image confidence (MMC), which leverages the complementary strengths of various multimodal medical images to assign a confidence measure to each voxel within the region of interest (ROI). MMC enables the generation of modality-specific ROI-enhanced images, providing a detailed depiction of both the boundaries and internal features of the ROI. By employing an interpretable mathematical model that propagates voxel confidence based on intervoxel correlations, MMC circumvents the need for model training, distinguishing it from deep learning-based methods. The alogorithm was evaluated qualitatively and quantitatively on 156 nasopharyngeal carcinoma cases and 1251 glioma cases. Qualitative assessments demonstrated MMC's accuracy in delineating lesion boundaries as well as capturing internal tumor characteristics. Quantitative analyses further revealed strong concordance between MMC and manual delineations. This study presents a cutting-edge algorithm for identifying and illustating ROI boundaries using multimodal 3D medical images. The versatility of the proposed method extends to both targets and organs at risk across various anatomic sites and multiple image modalities, enhancing its potential for accurate delineation of critical structures andmany image-related tasks in radiaton therapy and other fields.

Applications of MR Finger printing derived T1 and T2 values in Adult brain: A Systematic review

Introduction

Magnetic resonance imaging (MRI) is essential for brain imaging, but conventional methods rely on qualitative contrast, are time-intensive, and prone to variability. Magnetic resonance finger printing (MRF) addresses these limitations by enabling fast, simultaneous mapping of multiple tissue properties like T1, T2. Using dynamic acquisition parameters and a precomputed signal dictionary, MRF provides robust, qualitative maps, improving diagnostic precision and expanding clinical and research applications in brain imaging.

Methods

Database searches were performed through PubMed, Embase, Scopus, Web of science to identify relevant articles focusing on the application of MR finger printing in the adult brain. We utilized the preferred reporting items for systematic reviews and meta-analysis guidelines to extract data from the selected studies.

Results

Nine articles were included in the final review, with a total sample size of 332 participants. In healthy brains, notable regional, sex, age, and hemispheric variations were identified, particularly in the corpus callosum and thalamus. MRF effectively differentiated meningioma subtypes, glioma grades, and IDH mutation status, with T2 values providing particularly predictive for glioma classification. In brain metastases, significant relaxometry differences were noted between normal and lesional tissues. For multiple sclerosis, MRF values correlated with clinical and disability measures, distinguishing relapsing-remitting secondary progressive forms. In traumatic brain injury, longitudinal T1 changes strongly correlated with clinical recovery, surpassing T2 values.

Conclusions

The systematic review highlighted MRD as a groundbreaking technique that enhances neurological diagnosis by simultaneously quantifying T1 and T2 relaxation times. With reduced acquisition times, MRF outperforms conventional MRI in detecting subtle pathologies, distinguishing properties, and providing reliable biomarkers.

The Neural Frontier of Future Medical Imaging: A Review of Deep Learning for Brain Tumor Detection

Brain tumor detection is crucial in medical research due to high mortality rates and treatment challenges. Early and accurate diagnosis is vital for improving patient outcomes, however, traditional methods, such as manual Magnetic Resonance Imaging (MRI) analysis, are often time-consuming and error-prone. The rise of deep learning has led to advanced models for automated brain tumor feature extraction, segmentation, and classification. Despite these advancements, comprehensive reviews synthesizing recent findings remain scarce. By analyzing over 100 research papers over past half-decade (2019-2024), this review fills that gap, exploring the latest methods and paradigms, summarizing key concepts, challenges, datasets, and offering insights into future directions for brain tumor detection using deep learning. This review also incorporates an analysis of previous reviews and targets three main aspects: feature extraction, segmentation, and classification. The results revealed that research primarily focuses on Convolutional Neural Networks (CNNs) and their variants, with a strong emphasis on transfer learning using pre-trained models. Other methods, such as Generative Adversarial Networks (GANs) and Autoencoders, are used for feature extraction, while Recurrent Neural Networks (RNNs) are employed for time-sequence modeling. Some models integrate with Internet of Things (IoT) frameworks or federated learning for real-time diagnostics and privacy, often paired with optimization algorithms. However, the adoption of eXplainable AI (XAI) remains limited, despite its importance in building trust in medical diagnostics. Finally, this review outlines future opportunities, focusing on image quality, underexplored deep learning techniques, expanding datasets, and exploring deeper learning representations and model behavior such as recurrent expansion to advance medical imaging diagnostics.

Using magnetic resonance relaxometry to evaluate the safety and quality of induced pluripotent stem cell-derived spinal cord progenitor cells

BACKGROUND

The emergence of induced pluripotent stem cells (iPSCs) offers a promising approach for replacing damaged neurons and glial cells, particularly in spinal cord injuries (SCI). Despite its merits, iPSC differentiation into spinal cord progenitor cells (SCPCs) is variable, necessitating reliable assessment of differentiation and validation of cell quality and safety. Phenotyping is often performed via label-based methods including immunofluorescent staining or flow cytometry analysis. These approaches are often expensive, laborious, time-consuming, destructive, and severely limits their use in large scale cell therapy manufacturing settings. On the other hand, cellular biophysical properties have demonstrated a strong correlation to cell state, quality and functionality and can be measured with ingenious label-free technologies in a rapid and non-destructive manner.

METHOD

In this study, we report the use of Magnetic Resonance Relaxometry (MRR), a rapid and label-free method that indicates iron levels based on its readout (T2). Briefly, we differentiated human iPSCs into SCPCs and compared key iPSC and SCPC cellular markers to their intracellular iron content (Fe3+) at different stages of the differentiation process.

RESULTS

With MRR, we found that intracellular iron of iPSCs and SCPCs were distinctively different allowing us to accurately reflect varying levels of residual undifferentiated iPSCs (i.e., OCT4+ cells) in any given population of SCPCs. MRR was also able to predict Day 10 SCPC OCT4 levels from Day 1 undifferentiated iPSC T2 values and identified poorly differentiated SCPCs with lower T2, indicative of lower neural progenitor (SOX1) and stem cell (Nestin) marker expression levels. Lastly, MRR was able to provide predictive indications for the extent of differentiation to Day 28 spinal cord motor neurons (ISL-1/SMI-32) based on the T2 values of Day 10 SCPCs.

CONCLUSION

MRR measurements of iPSCs and SCPCs has clearly indicated its capabilities to identify and quantify key phenotypes of iPSCs and SCPCs for end-point validation of safety and quality parameters. Thus, our technology provides a rapid label-free method to determine critical quality attributes in iPSC-derived progenies and is ideally suited as a quality control tool in cell therapy manufacturing.

Inferring the genetic relationships between unsupervised deep learning-derived imaging phenotypes and glioblastoma through multi-omics approaches

This study aimed to investigate the genetic association between glioblastoma (GBM) and unsupervised deep learning-derived imaging phenotypes (UDIPs). We employed a combination of genome-wide association study (GWAS) data, single-nucleus RNA sequencing (snRNA-seq), and scPagwas (pathway-based polygenic regression framework) methods to explore the genetic links between UDIPs and GBM. Two-sample Mendelian randomization analyses were conducted to identify causal relationships between UDIPs and GBM. Colocalization analysis was performed to validate genetic associations, while scPagwas analysis was used to evaluate the relevance of key UDIPs to GBM at the cellular level. Among 512 UDIPs tested, 23 were found to have significant causal associations with GBM. Notably, UDIPs such as T1-33 (OR = 1.007, 95% CI = 1.001 to 1.012, P = .022), T1-34 (OR = 1.012, 95% CI = 1.001-1.023, P = .028), and T1-96 (OR = 1.009, 95% CI = 1.001-1.019, P = .046) were found to have a genetic association with GBM. Furthermore, T1-34 and T1-96 were significantly associated with GBM recurrence, with P-values < .0001 and P < .001, respectively. In addition, scPagwas analysis revealed that T1-33, T1-34, and T1-96 are distinctively linked to different GBM subtypes, with T1-33 showing strong associations with the neural progenitor-like subtype (NPC2), T1-34 with mesenchymal (MES2) and neural progenitor (NPC1) cells, and T1-96 with the NPC2 subtype. T1-33, T1-34, and T1-96 hold significant potential for predicting tumor recurrence and aiding in the development of personalized GBM treatment strategies.

Multimodal brain tumor segmentation and classification from MRI scans based on optimized DeepLabV3+ and interpreted networks information fusion empowered with explainable AI

Explainable artificial intelligence (XAI) aims to offer machine learning (ML) methods that enable people to comprehend, properly trust, and create more explainable models. In medical imaging, XAI has been adopted to interpret deep learning black box models to demonstrate the trustworthiness of machine decisions and predictions. In this work, we proposed a deep learning and explainable AI-based framework for segmenting and classifying brain tumors. The proposed framework consists of two parts. The first part, encoder-decoder-based DeepLabv3+ architecture, is implemented with Bayesian Optimization (BO) based hyperparameter initialization. The different scales are performed, and features are extracted through the Atrous Spatial Pyramid Pooling (ASPP) technique. The extracted features are passed to the output layer for tumor segmentation. In the second part of the proposed framework, two customized models have been proposed named Inverted Residual Bottleneck 96 layers (IRB-96) and Inverted Residual Bottleneck Self-Attention (IRB-Self). Both models are trained on the selected brain tumor datasets and extracted features from the global average pooling and self-attention layers. Features are fused using a serial approach, and classification is performed. The BO-based hyperparameters optimization of the neural network classifiers is performed and the classification results have been optimized. An XAI method named LIME is implemented to check the interpretability of the proposed models. The experimental process of the proposed framework was performed on the Figshare dataset, and an average segmentation accuracy of 92.68 % and classification accuracy of 95.42 % were obtained, respectively. Compared with state-of-the-art techniques, the proposed framework shows improved accuracy.

Machine learning-based nomogram for distinguishing between supratentorial extraventricular ependymoma and supratentorial glioblastoma

Objective

To develop a machine learning-based nomogram for distinguishing between supratentorial extraventricular ependymoma (STEE) and supratentorial glioblastoma (GBM).

Methods

We conducted a retrospective analysis on MRI datasets obtained from 140 patients who were diagnosed with STEE (n=48) and GBM (n=92) from two institutions. Initially, we compared seven different machine learning algorithms to determine the most suitable signature (rad-score). Subsequently, univariate and multivariate logistic regression analyses were performed to identify significant clinical predictors that can differentiate between STEE and GBM. Finally, we developed a nomogram by visualizing the rad-score and clinical features for clinical evaluation.

Results

The TreeBagger (TB) outperformed the other six algorithms, yielding the best diagnostic efficacy in differentiating STEE from GBM, with area under the curve (AUC) values of 0.735 (95% CI: 0.625-0.845) and 0.796 (95% CI: 0.644-0.949) in the training set and test set. Furthermore, the nomogram incorporating both the rad-score and clinical variables demonstrated a robust predictive performance with an accuracy of 0.787 in the training set and 0.832 in the test set.

Conclusion

The nomogram could serve as a valuable tool for non-invasively discriminating between STEE and GBM.

Combination chemotherapy via poloxamer 188 surface-modified PLGA nanoparticles that traverse the blood-brain-barrier in a glioblastoma model

The effect of chemotherapy for anti-glioblastoma is limited due to insufficient drug delivery across the blood-brain-barrier. Poloxamer 188-coated nanoparticles can enhance the delivery of nanoparticles across the blood-brain-barrier. This study presents the design, preparation, and evaluation of a combination of PLGA nanoparticles (PLGA NPs) loaded with methotrexate (P-MTX NPs) and PLGA nanoparticles loaded with paclitaxel (P-PTX NPs), both of which were surface-modified with poloxamer188. Cranial tumors were induced by implanting C6 cells in a rat model and MRI demonstrated that the tumors were indistinguishable in the two rats with P-MTX NPs + P-PTX NPs treated groups. Brain PET scans exhibited a decreased brain-to-background ratio which could be attributed to the diminished metabolic tumor volume. The expression of Ki-67 as a poor prognosis factor, was significantly lower in P-MTX NPs + P-PTX NPs compared to the control. Furthermore, the biodistribution of PLGA NPs was determined by carbon quantum dots loaded into PLGA NPs (P-CQD NPs), and quantitative analysis of ex-vivo imaging of the dissected organs demonstrated that 17.2 ± 0.6% of the NPs were concentrated in the brain after 48 h. The findings highlight the efficacy of combination nanochemotherapy in glioblastoma treatment, indicating the need for further preclinical studies.

Auto-segmentation of Adult-Type Diffuse Gliomas: Comparison of Transfer Learning-Based Convolutional Neural Network Model vs. Radiologists

Segmentation of glioma is crucial for quantitative brain tumor assessment, to guide therapeutic research and clinical management, but very time-consuming. Fully automated tools for the segmentation of multi-sequence MRI are needed. We developed and pretrained a deep learning (DL) model using publicly available datasets A (n = 210) and B (n = 369) containing FLAIR, T2WI, and contrast-enhanced (CE)-T1WI. This was then fine-tuned with our institutional dataset (n = 197) containing ADC, T2WI, and CE-T1WI, manually annotated by radiologists, and split into training (n = 100) and testing (n = 97) sets. The Dice similarity coefficient (DSC) was used to compare model outputs and manual labels. A third independent radiologist assessed segmentation quality on a semi-quantitative 5-scale score. Differences in DSC between new and recurrent gliomas, and between uni or multifocal gliomas were analyzed using the Mann-Whitney test. Semi-quantitative analyses were compared using the chi-square test. We found that there was good agreement between segmentations from the fine-tuned DL model and ground truth manual segmentations (median DSC: 0.729, std-dev: 0.134). DSC was higher for newly diagnosed (0.807) than recurrent (0.698) (p < 0.001), and higher for unifocal (0.747) than multi-focal (0.613) cases (p = 0.001). Semi-quantitative scores of DL and manual segmentation were not significantly different (mean: 3.567 vs. 3.639; 93.8% vs. 97.9% scoring ≥ 3, p = 0.107). In conclusion, the proposed transfer learning DL performed similarly to human radiologists in glioma segmentation on both structural and ADC sequences. Further improvement in segmenting challenging postoperative and multifocal glioma cases is needed.

Multiparametric simultaneous hybrid 18F-fluorodeoxyglucose positron emission tomography/magnetic resonance imaging (18F-FDG PET/MRI) incorporating intratumoral and peritumoral regions for grading of glioma

Background

Preoperative grading gliomas is essential for therapeutic clinical decision-making. Current non-invasive imaging modality for glioma grading were primarily focused on magnetic resonance imaging (MRI) or positron emission tomography (PET) of the tumor region. However, these methods overlook the peritumoral region (PTR) of tumor and cannot take full advantage of the biological information derived from hybrid-imaging. Therefore, we aimed to combine multiparameter from hybrid 18F-fluorodeoxyglucose (18F-FDG) PET/MRI of the solid component and PTR were combined for differentiating high-grade glioma (HGG) from low-grade glioma (LGG).

Methods

A total of 76 patients with pathologically confirmed glioma (41 HGG and 35 LGG) who underwent simultaneous 18F-FDG PET, arterial spin labelling (ASL), and diffusion-weighted imaging (DWI) with hybrid PET/MRI were retrospectively enrolled. The relative maximum standardized uptake value (rSUVmax), relative cerebral blood flow (rCBF), and relative minimum apparent diffusion coefficient (rADCmin) for the solid component and PTR at different distances outside tumoral border were compared. Receiver operating characteristic (ROC) curves were applied to assess the grading performance. A nomogram for HGG prediction was constructed.

Results

HGGs displayed higher rSUVmax and rCBF but lower rADCmin in the solid component and 5 mm-adjacent PTR, lower rADCmin in 10 mm-adjacent PTR, and higher rCBF in 15- and 20-mm-adjacent PTR. rSUVmax in solid component performed best [area under the curve (AUC) =0.865] as a single parameter for grading. Combination of rSUVmax in the solid component and adjacent 20 mm performed better (AUC =0.881). Integration of all 3 indicators in the solid component and adjacent 20 mm performed the best (AUC =0.928). The nomogram including rSUVmax, rCBF, and rADCmin in the solid component and 5-mm-adjacent PTR predicted HGG with a concordance index (C-index) of 0.906.

Conclusions

Multiparametric 18F-FDG PET/MRI from the solid component and PTR performed excellently in differentiating HGGs from LGGs. It can be used as a non-invasive and effective tool for preoperative grade stratification of patients with glioma, and can be considered in clinical practice.

Topography and Radiological Variables as Ancillary Parameters for Evaluating Tissue Adherence, Hypothalamic-Pituitary Dysfunction, and Recurrence in Craniopharyngioma: An Integrated Multidisciplinary Overview

Craniopharyngiomas continue to present a challenge in clinical practice due to their heterogeneity and unpredictable adherence to vital neurovascular structures, particularly the hypothalamus. This results in different degrees of hypothalamus-pituitary axis dysfunction and a lack of uniform consensus and treatment guidelines regarding optimal management. MRI and CT are complementary techniques in the preoperative diagnostic phase, enabling the precise definition of craniopharyngioma size, shape, and consistency, as well as guiding classification into histopathological subtypes and topographical categories. Meanwhile, MRI plays a crucial role in the immediate postoperative period and follow-up stages by identifying treatment-related changes and residual tumors. This pictorial essay aims to provide an overview of the role of imaging in identifying variables indicative of the adherence degree to the hypothalamus, hypothalamic-pituitary dysfunction, the extent of surgical excision, and prognosis. For a more comprehensive assessment, we choose to distinguish the following two scenarios: (1) the initial diagnosis phase, where we primarily discuss the role of radiological variables predictive of adhesions to the surrounding neurovascular structures and axis dysfunction which may influence the choice of surgical resection; (2) the early post-treatment follow-up phase, where we discuss the interpretation of treatment-related changes that impact outcomes.

Revolutionizing Brain Tumor Care: Emerging Technologies and Strategies

Glioblastoma multiforme (GBM) is one of the most aggressive forms of brain tumor, characterized by a daunting prognosis with a life expectancy hovering around 12-16 months. Despite a century of relentless research, only a select few drugs have received approval for brain tumor treatment, largely due to the formidable barrier posed by the blood-brain barrier. The current standard of care involves a multifaceted approach combining surgery, irradiation, and chemotherapy. However, recurrence often occurs within months despite these interventions. The formidable challenges of drug delivery to the brain and overcoming therapeutic resistance have become focal points in the treatment of brain tumors and are deemed essential to overcoming tumor recurrence. In recent years, a promising wave of advanced treatments has emerged, offering a glimpse of hope to overcome the limitations of existing therapies. This review aims to highlight cutting-edge technologies in the current and ongoing stages of development, providing patients with valuable insights to guide their choices in brain tumor treatment.

The FKBP51s Splice Isoform Predicts Unfavorable Prognosis in Patients with Glioblastoma

The primary treatment for glioblastoma (GBM) is removing the tumor mass as defined by MRI. However, MRI has limited diagnostic and predictive value. Tumor-associated macrophages (TAM) are abundant in GBM tumor microenvironment (TME) and are found in peripheral blood (PB). FKBP51 expression, with its canonical and spliced isoforms, is constitutive in immune cells and aberrant in GBM. Spliced FKBP51s supports M2 polarization. To find an immunologic signature that combined with MRI could advance in diagnosis, we immunophenotyped the macrophages of TME and PB from 37 patients with GBM using FKBP51s and classical M1-M2 markers. We also determined the tumor levels of FKBP51s, PD-L1, and HLA-DR. Tumors expressing FKBP51s showed an increase in various M2 phenotypes and regulatory T cells in PB, indicating immunosuppression. Tumors expressing FKBP51s also activated STAT3 and were associated with reduced survival. Correlative studies with MRI and tumor/macrophages cocultures allowed to interpret TAMs. Tumor volume correlated with M1 infiltration of TME. Cocultures with spheroids produced M1 polarization, suggesting that M1 macrophages may infiltrate alongside cancer stem cells. Cocultures of adherent cells developed the M2 phenotype CD163/FKBP51s expressing pSTAT6, a transcription factor enabling migration and invasion. In patients with recurrences, increased counts of CD163/FKBP51s monocyte/macrophages in PB correlated with callosal infiltration and were accompanied by a concomitant decrease in TME-infiltrating M1 macrophages. PB PD-L1/FKBP51s connoted necrotic tumors. In conclusion, FKBP51s identifies a GBM subtype that significantly impairs the immune system. Moreover, FKBP51s marks PB macrophages associated with MRI features of glioma malignancy that can aid in patient monitoring.

SIGNIFICANCE

Our research suggests that by combining imaging with analysis of monocyte/macrophage subsets in patients with GBM, we can enhance our understanding of the disease and assist in its treatment. We discovered a similarity in the macrophage composition between the TME and PB, and through association with imaging, we could interpret macrophages. In addition, we identified a predictive biomarker that drew more attention to immune suppression of patients with GBM.

Artificial intelligence in neuro-oncology: advances and challenges in brain tumor diagnosis, prognosis, and precision treatment

This review delves into the most recent advancements in applying artificial intelligence (AI) within neuro-oncology, specifically emphasizing work on gliomas, a class of brain tumors that represent a significant global health issue. AI has brought transformative innovations to brain tumor management, utilizing imaging, histopathological, and genomic tools for efficient detection, categorization, outcome prediction, and treatment planning. Assessing its influence across all facets of malignant brain tumor management- diagnosis, prognosis, and therapy- AI models outperform human evaluations in terms of accuracy and specificity. Their ability to discern molecular aspects from imaging may reduce reliance on invasive diagnostics and may accelerate the time to molecular diagnoses. The review covers AI techniques, from classical machine learning to deep learning, highlighting current applications and challenges. Promising directions for future research include multimodal data integration, generative AI, large medical language models, precise tumor delineation and characterization, and addressing racial and gender disparities. Adaptive personalized treatment strategies are also emphasized for optimizing clinical outcomes. Ethical, legal, and social implications are discussed, advocating for transparency and fairness in AI integration for neuro-oncology and providing a holistic understanding of its transformative impact on patient care.

Marker-controlled watershed algorithm and fuzzy C-means clustering machine learning: automated segmentation of glioblastoma from MRI images in a case series

INTRODUCTION AND IMPORTANCE

Automated segmentation of glioblastoma multiforme (GBM) from MRI images is crucial for accurate diagnosis and treatment planning. This paper presents a new and innovative approach for automating the segmentation of GBM from MRI images using the marker-controlled watershed segmentation (MCWS) algorithm.

CASE PRESENTATION AND METHODS

The technique involves several image processing techniques, including adaptive thresholding, morphological filtering, gradient magnitude calculation, and regional maxima identification. The MCWS algorithm efficiently segments images based on local intensity structures using the watershed transform, and fuzzy c-means (FCM) clustering improves segmentation accuracy. The presented approach achieved improved segmentation accuracy in detecting and segmenting GBM tumours from axial T2-weighted (T2-w) MRI images, as demonstrated by the mean characteristics performance metrics for GBM segmentation (sensitivity: 0.9905, specificity: 0.9483, accuracy: 0.9508, precision: 0.5481, F_measure: 0.7052, and jaccard: 0.9340).

CLINICAL DISCUSSION

The results of this study underline the importance of reliable and accurate image segmentation for effective diagnosis and treatment planning of GBM tumours.

CONCLUSION

The MCWS technique provides an effective and efficient approach for the segmentation of challenging medical images.

Clinical evaluation of augmented reality-based 3D navigation system for brachial plexus tumor surgery

BACKGROUND

Augmented reality (AR), a form of 3D imaging technology, has been preliminarily applied in tumor surgery of the head and spine, both are rigid bodies. However, there is a lack of research evaluating the clinical value of AR in tumor surgery of the brachial plexus, a non-rigid body, where the anatomical position varies with patient posture.

METHODS

Prior to surgery in 8 patients diagnosed with brachial plexus tumors, conventional MRI scans were performed to obtain conventional 2D MRI images. The MRI data were then differentiated automatically and converted into AR-based 3D models. After point-to-point relocation and registration, the 3D models were projected onto the patient's body using a head-mounted display for navigation. To evaluate the clinical value of AR-based 3D models compared to the conventional 2D MRI images, 2 senior hand surgeons completed questionnaires on the evaluation of anatomical structures (tumor, arteries, veins, nerves, bones, and muscles), ranging from 1 (strongly disagree) to 5 (strongly agree).

RESULTS

Surgeons rated AR-based 3D models as superior to conventional MRI images for all anatomical structures, including tumors. Furthermore, AR-based 3D models were preferred for preoperative planning and intraoperative navigation, demonstrating their added value. The mean positional error between the 3D models and intraoperative findings was approximately 1 cm.

CONCLUSIONS

This study evaluated, for the first time, the clinical value of an AR-based 3D navigation system in preoperative planning and intraoperative navigation for brachial plexus tumor surgery. By providing more direct spatial visualization, compared with conventional 2D MRI images, this 3D navigation system significantly improved the clinical accuracy and safety of tumor surgery in non-rigid bodies.

Brain tumor classification using fine-tuned transfer learning models on magnetic resonance imaging (MRI) images

OBJECTIVE

Brain tumors are a leading global cause of mortality, often leading to reduced life expectancy and challenging recovery. Early detection significantly improves survival rates. This paper introduces an efficient deep learning model to expedite brain tumor detection through timely and accurate identification using magnetic resonance imaging images.

METHODS

Our approach leverages deep transfer learning with six transfer learning algorithms: VGG16, ResNet50, MobileNetV2, DenseNet201, EfficientNetB3, and InceptionV3. We optimize data preprocessing, upsample data through augmentation, and train the models using two optimizers: Adam and AdaMax. We perform three experiments with binary and multi-class datasets, fine-tuning parameters to reduce overfitting. Model effectiveness is analyzed using various performance scores with and without cross-validation.

RESULTS

With smaller datasets, the models achieve 100% accuracy in both training and testing without cross-validation. After applying cross-validation, the framework records an outstanding accuracy of 99.96% with a receiver operating characteristic of 100% on average across five tests. For larger datasets, accuracy ranges from 96.34% to 98.20% across different models. The methodology also demonstrates a small computation time, contributing to its reliability and speed.

CONCLUSION

The study establishes a new standard for brain tumor classification, surpassing existing methods in accuracy and efficiency. Our deep learning approach, incorporating advanced transfer learning algorithms and optimized data processing, provides a robust and rapid solution for brain tumor detection.

Comparative study between dynamic susceptibility contrast magnetic resonance imaging and arterial spin labelling perfusion in differentiating low-grade from high-grade brain tumours

Purpose

Our aim was to distinguish between low-grade and high-grade brain tumours on the basis of dynamic susceptibility contrast (DSC) magnetic resonance imaging (MRI) perfusion and arterial spin labelling (ASL) perfusion and to compare DSC and ASL techniques.

Material and methods

Forty-one patients with brain tumours were evaluated by 3-Tesla MRI. Conventional and perfusion MRI imaging with a 3D pseudo-continuous ASL (PCASL) and DSC perfusion maps were evaluated. Three ROIs were placed to obtain cerebral blood value (CBV) and cerebral blood flow (CBF) in areas of maximum perfusion in brain tumour and normal grey matter. Histopathological diagnosis was considered as the reference. ROC analysis was performed to compare the diagnostic performance and to obtain a feasible cut-off value of perfusion parameters to differentiate low-grade and high-grade brain tumours.

Results

Normalised perfusion parameters with grey matter (rCBF or rCBV lesion/NGM) of malignant lesions were significantly higher than those of benign lesions in both DSC (normalised rCBF of 2.16 and normalised rCBV of 2.63) and ASL (normalised rCBF of 2.22) perfusion imaging. The normalised cut-off values of DSC (rCBF of 1.1 and rCBV of 1.4) and ASL (rCBF of 1.3) showed similar specificity and near similar sensitivity in distinguishing low-grade and high-grade brain tumours.

Conclusions

Quantitative analysis of perfusion parameters obtained by both DSC and ASL perfusion techniques can be reliably used to distinguish low-grade and high-grade brain tumours. Normalisation of these values by grey matter gives us more reliable parameters, eliminating the different technical parameters involved in both the techniques.

Magnetic resonance imaging-derived parameters to predict response to regorafenib in recurrent glioblastoma

PURPOSE

Regorafenib is a multikinase inhibitor, approved as a preferred regimen for recurrent glioblastoma (rGB). Although its effects on prolonging survival could seem modest, it is still unclear whether a subset of patients, potentially identifiable by imaging biomarkers, might experience a more substantial positive effect. Our aim was to evaluate the potential value of magnetic resonance imaging-derived parameters as non-invasive biomarkers to predict response to regorafenib in patients with rGB.

METHODS

20 patients with rGB underwent conventional and advanced MRI at diagnosis (before surgery), at recurrence and at first follow-up (3 months) during regorafenib. Maximum relative cerebral blood volume (rCBVmax) value, intra-tumoral susceptibility signals (ITSS), apparent diffusion coefficient (ADC) values, and contrast-enhancing tumor volumes were tested for correlation with response to treatment, progression-free survival (PFS), and overall survival (OS). Response at first follow-up was assessed according to Response Assessment in Neuro-Oncology (RANO) criteria.

RESULTS

8/20 patients showed stable disease at first follow-up. rCBVmax values of the primary glioblastoma (before surgery) significantly correlated to treatment response; specifically, patients with stable disease displayed higher rCBVmax compared to progressive disease (p = 0.04, 2-group t test). Moreover, patients with stable disease showed longer PFS (p = 0.02, 2-group t test) and OS (p = 0.04, 2-group t test). ITSS, ADC values, and contrast-enhancing tumor volumes showed no correlation with treatment response, PFS nor OS.

CONCLUSION

Our results suggest that rCBVmax of the glioblastoma at diagnosis could serve as a non-invasive biomarker of treatment response to regorafenib in patients with rGB.

Predicting the WHO Grading of Pediatric Brain Tumors Based on Their MRI Appearance: A Retrospective Study

The treatment of central nervous system (CNS) tumors constitutes a significant part of a pediatric neurosurgeon's workload. The classification of such neoplasms spans many entities. These include low- and high-grade lesions, with both occurring in the population of patients under 18 years of age. Magnetic resonance imaging serves as the imaging method of choice for neoplastic lesions of the brain. Through its different modalities, such as T1, T2, T1 C+, apparent diffusion coefficient (ADC), diffusion-weighted imaging (DWI), susceptibility-weighted imaging (SWI), fluid-attenuated inversion recovery (FLAIR), etc., it allows the medical team to plan the therapeutic process accordingly while also possibly suggesting the specific tumor subtype prior to obtaining a definitive histological diagnosis. We conducted a retrospective study spanning 32 children treated surgically for brain tumors between July 2021 and January 2023 who had a precise histological diagnosis determined by using the 2021 WHO Classification of Tumors of the Central Nervous System. We divided them into two groups (high-grade and low-grade tumors, i.e., WHO grades 1 and 2, and grades 3 and 4, respectively) and analyzed their demographic data and preoperative MRI results. This was done using the following criteria: sub or supratentorial location of the tumor; lesion is circumscribed or infiltrating; solid, cystic, or mixed solid and cystic character of the tumor; number of compartments in cystic lesions; signal intensity (hypo-, iso-, hyperintensity sequences: T1, T2, T1 C+); presence of restricted diffusion; the largest diameter of the solid component and/or the largest diameter of the largest cyst in the transverse section. Then, we examined the results to find any correlation between the lesions' morphologies and their final assigned degree of malignancy. We found that the only radiological criteria correlating with the final WHO grade of the tumor were an infiltrative pattern of growth (25% of low-grade lesions, 75% of high-grade; p = 0.006) and the presence of a cystic component in the tumor (in 68.75% of low-grade tumors and 43.75% of high-grade tumors; p = 0.041). The only other feature close to attaining statistical significance was diffusion restriction (33.3% of low-grade tumors, 66.7% high-grade; p = 0.055). Older children tended to present with tumors of lower degrees of malignancy, and there was a predominance of female patients (21 female, 11 male).

Advanced Magnetic Resonance Imaging in the Evaluation of Treated Glioblastoma: A Pictorial Essay

MRI plays a key role in the evaluation of post-treatment changes, both in the immediate post-operative period and during follow-up. There are many different treatment's lines and many different neuroradiological findings according to the treatment chosen and the clinical timepoint at which MRI is performed. Structural MRI is often insufficient to correctly interpret and define treatment-related changes. For that, advanced MRI modalities, including perfusion and permeability imaging, diffusion tensor imaging, and magnetic resonance spectroscopy, are increasingly utilized in clinical practice to characterize treatment effects more comprehensively. This article aims to provide an overview of the role of advanced MRI modalities in the evaluation of treated glioblastomas. For a didactic purpose, we choose to divide the treatment history in three main timepoints: post-surgery, during Stupp (first-line treatment) and at recurrence (second-line treatment). For each, a brief introduction, a temporal subdivision (when useful) or a specific drug-related paragraph were provided. Finally, the current trends and application of radiomics and artificial intelligence (AI) in the evaluation of treated GB have been outlined.