Brain Magnetic Resonance Imaging Images Research Articles

An investigation into the applicability of rapid artificial intelligence‐assisted compressed sensing in brain magnetic resonance imaging performed at 5 Tesla field strength

AbstractBackgroundBrain magnetic resonance imaging (MRI) at 5 T offers unprecedented spatial resolution but is often limited by long scan times. Acceleration techniques, such as compressed sensing (CS) and artificial intelligence‐assisted compressed sensing (ACS), have the potential to speed up the acquisition process while maintaining image quality. This study aims to evaluate and compare the performance of CS and ACS (with various acceleration factors) in brain MRI imaging at 5 T.MethodsIn this study, we enrolled 12 healthy volunteers and compared ACS‐accelerated 5 T brain MRI with conventional methods of CS. The ACS acceleration factors for the brain protocol, consisting of 3D T1‐weighted sequences and 2D T2‐weighted sequences, were optimized in a pilot study on healthy volunteers (acceleration factor, 2.06–3.41× in T2‐weighted imaging and 3.52–8.49× in T1‐weighted imaging). We evaluated the images acquired from patients using various acceleration methods on the basis of acquisition times, the signal‐to‐noise ratio (SNR), the contrast‐to‐noise ratio, subjective image quality, and diagnostic agreement.ResultsOur findings revealed that ACS acceleration significantly reduced the acquisition times for T1‐ and T2‐weighted sequences by up to 43% and 53%, respectively, compared with traditional CS at 5 T. Importantly, this acceleration was achieved while maintaining excellent image quality, demonstrated by higher or comparable SNR and contrast‐to‐noise ratio values.ConclusionsThe optimal ACS acceleration factors for 5 T brain MRI were determined to be 2.73× for 2D T2‐weighted sequences and 6.5× for 3D T1‐weighted sequences. ACS not only facilitates rapid imaging but also ensures comparable image quality and diagnostic performance, highlighting its potential to revolutionize high‐field MRI scanning.

A Parkinson's disease-related nuclei segmentation network based on CNN-Transformer interleaved encoder with feature fusion.

Automatic segmentation of Parkinson's disease (PD) related deep gray matter (DGM) nuclei based on brain magnetic resonance imaging (MRI) is significant in assisting the diagnosis of PD. However, due to the degenerative-induced changes in appearance, low tissue contrast, and tiny DGM nuclei size in elders' brain MRI images, many existing segmentation models are limited in the application. To address these challenges, this paper proposes a PD-related DGM nuclei segmentation network to provide precise prior knowledge for aiding diagnosis PD. The encoder of network is designed as an alternating encoding structure where the convolutional neural network (CNN) captures spatial and depth texture features, while the Transformer complements global position information between DGM nuclei. Moreover, we propose a cascaded channel-spatial-wise block to fuse features extracted by the CNN and Transformer, thereby achieving more precise DGM nuclei segmentation. The decoder incorporates a symmetrical boundary attention module, leveraging the symmetrical structures of bilateral nuclei regions by constructing signed distance maps for symmetric differences, which optimizes segmentation boundaries. Furthermore, we employ a dynamic adaptive region of interests weighted Dice loss to enhance sensitivity towards smaller structures, thereby improving segmentation accuracy. In qualitative analysis, our method achieved optimal average values for PD-related DGM nuclei (DSC: 0.854, IOU: 0.750, HD95: 1.691 mm, ASD: 0.195 mm). Experiments conducted on multi-center clinical datasets and public datasets demonstrate the good generalizability of the proposed method. Furthermore, a volumetric analysis of segmentation results reveals significant differences between HCs and PDs. Our method holds promise for assisting clinicians in the rapid and accurate diagnosis of PD, offering a practical method for the imaging analysis of neurodegenerative diseases.

Hybrid improved fuzzy C-means and watershed segmentation to classify Alzheimer’s using deep learning

Brain damage and deficits in interactions among brain cells are the primary causes of dementia and Alzheimer’s disease (AD). Despite ongoing research, no effective medications have yet been developed for these conditions. Therefore, early detection is crucial for managing the progression of these disorders. In this study, we introduce a novel tool for detecting AD using non invasive medical tests, such as magnetic resonance imaging (MRI). Our method employs fuzzy C-means clustering to identify features that enhance image accuracy. The standard fuzzy C-means algorithm has been augmented with fuzzy components to improve clustering performance. This enhanced approach optimizes segmentation by extracting image information and utilizing a sliding window to calculate center coordinates and establish a stable group matrix. These critical features are subsequently integrated with a two-phase watershed segmentation process. The resulting segmented images are then used to train an optimal convolutional neural network (CNN) for AD classification. Our methodology demonstrated a 98.20% accuracy rate in the detection and classification of segmented MRI brain images, highlighting its efficacy in identifying disease types.

Improving brain MRI denoising using convolutional AutoEncoder and sparse representations

Magnetic Resonance Imaging (MRI) is an essential tool for diagnosing and monitoring diseases under various conditions. However, noise often degrades image quality, leading to inaccurate diagnoses. To address this issue, a Convolutional AutoEncoder-based Orthogonal Matching Pursuit (CAE-OMP) model is proposed for brain MRI image denoising. In this model, the encoder block extracts relevant features from the image while reducing overfitting, and the OMP algorithm creates a sparse representation to enhance denoising. To improve performance and computational efficiency, the traditional greedy search process in OMP is replaced with the Crossover Boosted Elephant Herd Optimization (CBEHO) algorithm. Rather than searching for atoms, CBEHO optimizes parameter selection, thereby reducing search times and enhancing convergence in the OMP process. Using this optimized sparse representation, the model iteratively improves the original image’s approximation by updating residuals and the support set. The decoder block then reconstructs the denoised image features. The proposed method was tested on multiple datasets, including the RSNA MICCAI PNG dataset, the Brain Tumor Detection MRI (BTD-MRI) dataset, the Brain Tumor Classification MRI (BTC-MRI) Images dataset, and the Brain Tumor Segmentation (BraTS2020) dataset. The results show that the CAE-OMP model achieves Structural Similarity Index (SSIM) and Peak Signal-to-Noise Ratio (PSNR) values of 0.989 and 47.345 on the BTD-MRI dataset, 0.985 and 46.321 on the RSNA MICCAI PNG dataset, 0.978 and 45.453 on the BTC-MRI dataset, and 0.981 and 46.892 on the BraTS2020 dataset, all evaluated at a 15% noise level. These outcomes indicate that the proposed CAE-OMP model outperforms existing methods, demonstrating superior efficiency for denoising brain MRI images.

Large sample size MRI measurements to assess the relation between cardiac function and structure and white matter hyperintensity volumes

Abstract Background People with established cardiovascular disease (CVD) are at risk of early cognitive decline, and neurological diseases such as dementia. CVD and neurological diseases share common risk factors, such as blood pressure, sedentary lifestyle, as well as genomic risk factors such as mutations in APOE4 [1]. It's unclear to what extent the co-occurrence of CVD and neurological diseases is explained by these risk factors, or whether there is a direct association where CVD is a first step towards poorer neurological health. White matter hyperintensities (WMH) of presumed vascular origin are damaged areas in the brain observed through magnetic resonance imaging (MRI). These lesions are associated with progressive neurological disease such as vascular dementia, as well with risk factors for CVD [2]. Purpose We set out to determine to what extent CMR measurements associated with WMH independent of known cerebrovascular risk factors. Methods Cardiac and brain MRI images were analysed for 23,963 UK biobank (UKB) participants. WMH analysis included 5 brain regions (Frontal[F], Parietal[P], Temporal[T], Occipital[O], Basal Ganglia and Thalami [BGT]) and total brain WMH volume. Cardiac traits included stroke volumes, atrial volumes, ejection fractions of right and left chambers, left ventricular (LV) strain, LV wall thickness and aortic areas. Multivariable regression analysis was carried out, where each cardiac trait was regressed on each brain region and adjusted for: age, difference between age at initial visit and imaging visit, sex, systolic and diastolic blood pressure, Hba1c, household income, educational status, Townsend score and family history of disease (Alzheimer’s, Parkinson's, high blood pressure, stroke, heart disease) . Results were evaluated against a multiple testing correct p-value threshold of 0.05, reflecting the number of principle components(n=10) necessary to explain 90% of the cardiac trait variability. Results For 5 cardiac traits (right ventricular stroke volume, left atrial stroke volume, left ventricular stroke volume, left ventricular ejection fraction and right ventricular end diastolic volume) higher values were associated with lower WMH volumes in all brain regions. Additionally, some cardiac traits such as LV cardiac output associated with specific brain regions (T, O, BGT). Additionally, APOE- ε4 stratified analyses indicated, homozygous carriers show a greater association between higher minimum LA and RA volumes and higher WMH volumes in the occipital lobe. Conclusions This study gives an insight into the WMH burden in a large population of adults. It highlights the associations of cardiac traits with white matter hyperintensity volumes in different brain regions independent of known risk factors. Using cardiac traits as surrogate markers of different cardiac disease could explain how cardiac functionality defines WMH volume distribution and subsequently the wider relationship between CVD and cerebrovascular disease.Figure 1Graphical Abstract

Brain tumor classification utilizing pixel distribution and spatial dependencies higher-order statistical measurements through explainable ML models

Brain tumors are among the most fatal and devastating diseases, and they often result in a significant reduction in life expectancy. The devising of treatment plans that can extend the lives of affected individuals hinges on an accurate diagnosis of these tumors. Identifying and analyzing large volumes of magnetic resonance imaging (MRI) data manually proves to be both challenging and time-consuming. As a result, there exists a pressing need for a reliable machine-learning approach to accurately diagnose brain tumors, and numerous methods have already been proposed over the last decade. In this paper, a novel, comprehensive approach is proposed for identifying and classifying a given MR brain image as abnormal. Three common brain diseases, namely glioma, meningioma, and pituitary tumor, are chosen as abnormal brains, and the Figshare MRI brain image dataset was collected from the Kaggle and IEEE websites. The proposed method is initiated by employing 1st-order statistics, 2nd-order statistics, and higher-order transformed (DWT) feature extraction to extract features from images. Then missing data is addressed and handled using KNNImputer, followed by the application of the ExtratreesClassifier and PCA feature selection methods to identify the most relevant features and reduce the dimensions of these features. Subsequently, the reduced features are submitted to seven machine learning models, namely RF, GB, CB, SVM, LGBM, DT, and LR. The strategy of k-fold cross-validation is utilized to enhance the performance of those models. Finally, the models are evaluated using XAI approaches, which ensure transparent decision-making processes and provide insights into the model’s predictions. Remarkably, our approach achieves the highest accuracy, precision, recall, F1 score, MCC, Kappa, AUC-ROC, and R2, as well as the lowest loss, among the seven models evaluated, proving its effectiveness and applicability in multiple analytic applications relying on publicly available datasets.

Robust brain MRI image classification with SIBOW-SVM

Primary Central Nervous System tumors in the brain are among the most aggressive diseases affecting humans. Early detection and classification of brain tumor types, whether benign or malignant, glial or non-glial, is critical for cancer prevention and treatment, ultimately improving human life expectancy. Magnetic Resonance Imaging (MRI) is the most effective technique for brain tumor detection, generating comprehensive brain scans. However, human examination can be error-prone and inefficient due to the complexity, size, and location variability of brain tumors. Recently, automated classification techniques using machine learning methods, such as Convolutional Neural Networks (CNNs), have demonstrated significantly higher accuracy than manual screening. However, deep learning-based image classification methods, including CNNs, face challenges in estimating class probabilities without proper model calibration (Guo et al., 2017; Minderer et al., 2021). In this paper, we propose a novel brain tumor image classification method called SIBOW-SVM, which integrates the Bag-of-Features model with SIFT feature extraction and weighted Support Vector Machines. This new approach can effectively extract hidden image features, enabling differentiation of various tumor types, provide accurate label predictions, and estimate probabilities of images belonging to each class, offering high-confidence classification decisions. We have also developed scalable and parallelable algorithms to facilitate the practical implementation of SIBOW-SVM for massive image datasets. To benchmark our method, we apply SIBOW-SVM to a public dataset of brain tumor MRI images containing four classes: glioma, meningioma, pituitary, and normal. Our results demonstrate that the new method outperforms state-of-the-art techniques, including CNNs, in terms of uncertainty quantification, classification accuracy, computational efficiency, and data robustness.

Computer Aided Based Performance Analysis of Glioblastoma Tumor Detection Methods using UNET-CNN

Brain tumors are the life killing and threatening disease which affects all age groups around the world. The timely detection and followed by the perspective treatments saves the human life. The tumor regions in brain are detected and segmented using UNET-CNN architecture in this paper. During training process of the proposed work, both Glioblastoma and Healthy brain Magnetic Resonance Imaging (MRI) is preprocessed and then multi level transform is applied on the preprocessed image. The features are further computed from the transformed coefficients and these features are trained by UNET-CNN architecture to obtain trained vectors. During testing process of the proposed work, the test brain MRI image is preprocessed and then decomposed coefficients are obtained by multi level transform. Features are computed from these decomposed coefficients and they are classified using UNET-CNN architecture with the trained vectors. The simulation results of the developed methodology are compared with similar studies on both BRATS 2017 and BRATS 2018 datasets

A novel classification framework for genome-wide association study of whole brain MRI images using deep learning.

Genome-wide association studies (GWASs) have been widely applied in the neuroimaging field to discover genetic variants associated with brain-related traits. So far, almost all GWASs conducted in neuroimaging genetics are performed on univariate quantitative features summarized from brain images. On the other hand, powerful deep learning technologies have dramatically improved our ability to classify images. In this study, we proposed and implemented a novel machine learning strategy for systematically identifying genetic variants that lead to detectable nuances on Magnetic Resonance Images (MRI). For a specific single nucleotide polymorphism (SNP), if MRI images labeled by genotypes of this SNP can be reliably distinguished using machine learning, we then hypothesized that this SNP is likely to be associated with brain anatomy or function which is manifested in MRI brain images. We applied this strategy to a catalog of MRI image and genotype data collected by the Alzheimer's Disease Neuroimaging Initiative (ADNI) consortium. From the results, we identified novel variants that show strong association to brain phenotypes.

ADVANCEMENTS IN ALZHEIMER’S DIAGNOSIS THROUGH MRI USING BAYESIAN CONVOLUTIONAL NEURAL NETWORKS AND VARIATIONAL INFERENCE

Alzheimer’s disease is one of the brain disorders that can be deadly in older. The disease is less treated and less recognized, but Alzheimer’s disease is now a significant public health problem. Early detection of the disease can significantly reduce symptoms. However, the lack of medical personnel makes handling this disease complex. Therefore, an automatic diagnosis of Alzheimer’s disease is needed with a Magnetic Resonance Imaging (MRI) examination to get an accurate diagnosis of the disease. This study classified the type of Alzheimer’s disease with deep learning methods using the Bayesian Convolutional Neural Network (BCNN) and the Variational Inference (VI) technique. It aims to determine image classification and accuracy level at the level of Alzheimer’s disease by using 2,400 brain MRI images, divided into three classes (non-demented, very mild demented, and mild demented) based on severity. The data was acquired from the kaggle.com website. We use a dataset scenario of 80% for training and 20% for testing, 100x100 pixels, kernel size 3x3, and optimizer Adam with epoch 200. The accuracy of the image classification process is 80%. The non-demented label predicts that the uncertainty is 0.371, and the other uncertainty prediction is 0.002. The ability to anticipate uncertainty enables clinicians to make informed decisions regarding the reliability of the model’s output and the need for additional validation or confirmation.

An optimised deep learning approach for alzheimer’s disease classification

<p>Alzheimer’s disease (AD) is a progressive and incurable brain disorder. It starts out subtly and gets worse with time. 60 to 70 percent of dementia cases are brought on by this illness. An Alzheimer’s patient is diagnosed every two seconds, according to research. The complexity of the brain makes it often very challenging to identify in elderly people. In the area of medical imaging, deep learning is growing. Several deep learning techniques that attempted to identify and categorise the magnetic resonance imaging (MRI) brain images into four stages of AD will be compared in this work. 6400 MRI brain images were extracted from a dataset and divided into training, validation, and testing datasets. In our research on twelve deep learning architectures, inceptionV3 has given the best results with 99.56% and 97.75% accuracy on train and validation, respectively, and on test data, the model has achieved an accuracy of 95.81%. We trained the models using optimised ImageNet weights, which resulted in higher accuracy across all twelve models.</p>

Mask region-based convolutional neural network and VGG-16 inspired brain tumor segmentation

The process of brain tumour segmentation entails locating the tumour precisely in images. Magnetic Resonance Imaging (MRI) is typically used by doctors to find any brain tumours or tissue abnormalities. With the use of region-based Convolutional Neural Network (R-CNN) masks, Grad-CAM and transfer learning, this work offers an effective method for the detection of brain tumours. Helping doctors make extremely accurate diagnoses is the goal. A transfer learning-based model has been suggested that offers high sensitivity and accuracy scores for brain tumour detection when segmentation is done using R-CNN masks. To train the model, the Inception V3, VGG-16, and ResNet-50 architectures were utilised. The Brain MRI Images for Brain Tumour Detection dataset was utilised to develop this method. This work's performance is evaluated and reported in terms of recall, specificity, sensitivity, accuracy, precision, and F1 score. A thorough analysis has been done comparing the proposed model operating with three distinct architectures: VGG-16, Inception V3, and Resnet-50. Comparing the proposed model, which was influenced by the VGG-16, to related works also revealed its performance. Achieving high sensitivity and accuracy percentages was the main goal. Using this approach, an accuracy and sensitivity of around 99% were obtained, which was much greater than current efforts.

Brain Tumor Classification Using Vanilla Convolutional Neural Networks

Brain tumours are a common and dangerous type of malignant tumour that, if not detected early enough, can cut short a patient's life. The segmentation and classification of brain tumours using solely traditional medical image processing is a difficult and time-consuming task. Various imaging modalities, such as Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and ultrasound image, are frequently utilized to assess brain, lung, liver, breast, prostate and other tumours. MRI images are specifically utilised in these analyses to detect brain tumours. As a result, developing approaches for detecting, recognizing, and classifying the conditions based on image analysis becomes essential. A comprehensive and automatic classification system is important to saving human lives. The geographical and anatomical heterogeneity of brain tumours makes automatic categorization challenging. This study proposes an automated method for detecting brain tumours using Convolutional Neural Networks (CNNs) classification, with the primary goal of developing a deep learning model that is capable of accurately identifying and classifying images as either having a brain tumour or not. In this paper, we provide a classification model for brain tumours based on a Deep Convolutional Neural Network with a vanilla neural network technique. The proposed method's performance on a publicly available dataset of 3000 Brain MRI Images yielded superior results, with accuracy and F1 score of 98.00 percent and 98.00 percent, respectively. This study shows that the proposed vanilla-CNN model can be used to make it easier for brain tumours to be automatically classified

Innovative Approaches for Early Alzheimer’s Disease Detection through Novel Analysis of Brain MRI Images

Alzheimer's is one of the most well-known reasons for Dementia. Alzheimer's sickness (AD) is an irreversible, moderate cerebrum issue that gradually obliterates memory and thinking abilities. Alzheimer's and different types of dementia positioned seventh driving reason for death by WHO. Picture handling is broadly utilized in the clinical field to recognize illness and help specialists in dynamics dependent on perception. The paper means to recognize the Alzheimer's sickness at the most punctual with the goal that patients can be forestalled before irreversible changes happen in the mind. We propose the picture handling method to deal with the Magnetic Resonance Imaging (MRI) of the cerebrum from the pivotal plane, coronal plane, and sagittal plane. The picture division is utilized to feature the impacted locale in cerebrum MRI. The analyzed area in mind MRI incorporates the hippocampus and volume of the cerebrum. A similar ID of people impacted with the Alzheimer's illness, Healthy partner, and Mild Cognitive weakness is finished.

Brain magnetic resonance image (MRI) segmentation using multimodal optimization

AbstractOne of the highly focused areas in the medical science community is segmenting tumors from brain magnetic resonance imaging (MRI). The diagnosis of malignant tumors at an early stage is necessary to provide treatment for patients. The patient’s prognosis will improve if it is detected early. Medical experts use a manual method of segmentation when making a diagnosis of brain tumors. This study proposes a new approach to simplify and automate this process. In recent research, multi-level segmentation has been widely used in medical image analysis, and the effectiveness and precision of the segmentation method are directly tied to the number of segments used. However, choosing the appropriate number of segments is often left up to the user and is challenging for many segmentation algorithms. The proposed method is a modified version of the 3D Histogram-based segmentation method, which can automatically determine an appropriate number of segments. The general algorithm contains three main steps: The first step is to use a Gaussian filter to smooth the 3D RGB histogram of an image. This eliminates unreliable and non-dominating histogram peaks that are too close together. Next, a multimodal particle swarm optimization method identifies the histogram’s peaks. In the end, pixels are placed in the cluster that best fits their characteristics based on the non-Euclidean distance. The proposed algorithm has been applied to a Cancer Imaging Archive (TCIA) and brain MRI Images for brain Tumor detection dataset. The results of the proposed method are compared with those of three clustering methods: FCM, FCM_FWCW, and FCM_FW. In the comparative analysis of the three algorithms across various MRI slices. Our algorithm consistently demonstrates superior performance. It achieves the top mean rank in all three metrics, indicating its robustness and effectiveness in clustering. The proposed method is effective in experiments, proving its capacity to find the proper clusters.

Improved neurological diagnoses and treatment strategies via automated human brain tissue segmentation from clinical magnetic resonance imaging

ObjectiveSegmentation of medical images is a crucial process in various image analysis applications. Automated segmentation methods excel in accuracy when compared to manual segmentation in the context of medical image analysis. One of the essential phases in the quantitative analysis of the brain is automated brain tissue segmentation using clinically obtained magnetic resonance imaging (MRI) data. It allows for precise quantitative examination of the brain, which aids in diagnosis, identification, and classification of disorders. Consequently, the efficacy of the segmentation approach is crucial to disease diagnosis and treatment planning. MethodsThis study presented a hybrid optimization method for segmenting brain tissue in clinical MRI scans using a fractional Henry horse herd gas optimization-based Shepard convolutional neural network (FrHHGO-based ShCNN). To segment the clinical brain MRI images into white matter (WM), grey matter (GM), and cerebrospinal fluid (CSF) tissues, the proposed framework was evaluated on the Lifespan Human Connectome Projects (HCP) database. The hybrid optimization algorithm, FrHHGO, integrates the fractional Henry gas optimization (FHGO) and horse herd optimization (HHO) algorithms. Training required 30 min, whereas testing and segmentation of brain tissues from an unseen image required an average of 12 s. ResultsCompared to the results obtained with no refinements, the Skull stripping refinement showed significant improvement. As the method included a preprocessing stage, it was flexible enough to enhance image quality, allowing for better results even with low-resolution input. Maximum precision of 93.2%, recall of 91.5%, Dice score of 91.1%, and F1-score of 90.5% were achieved using the proposed FrHHGO-based ShCNN, which was superior to all other approaches. ConclusionThe proposed method may outperform existing state-of-the-art methodologies in qualitative and quantitative measurements across a wide range of medical modalities. It might demonstrate its potential for real-life clinical application.

Sheep (Ovis aries) training protocol for voluntary awake and unrestrained structural brain MRI acquisitions

Magnetic resonance imaging (MRI) is a non-invasive technique that requires the participant to be completely motionless. To date, MRI in awake and unrestrained animals has only been achieved with humans and dogs. For other species, alternative techniques such as anesthesia, restraint and/or sedation have been necessary. Anatomical and functional MRI studies with sheep have only been conducted under general anesthesia. This ensures the absence of movement and allows relatively long MRI experiments but it removes the non-invasive nature of the MRI technique (i.e., IV injections, intubation). Anesthesia can also be detrimental to health, disrupt neurovascular coupling, and does not permit the study of higher-level cognition. Here, we present a proof-of-concept that sheep can be trained to perform a series of tasks, enabling them to voluntarily participate in MRI sessions without anesthesia or restraint. We describe a step-by-step training protocol based on positive reinforcement (food and praise) that could be used as a basis for future neuroimaging research in sheep. This protocol details the two successive phases required for sheep to successfully achieve MRI acquisitions of their brain. By providing structural brain MRI images from six out of ten sheep, we demonstrate the feasibility of our training protocol. This innovative training protocol paves the way for the possibility of conducting animal welfare-friendly functional MRI studies with sheep to investigate ovine cognition.

Enhancing brain tumor classification in MRI scans with a multi-layer customized convolutional neural network approach.

The necessity of prompt and accurate brain tumor diagnosis is unquestionable for optimizing treatment strategies and patient prognoses. Traditional reliance on Magnetic Resonance Imaging (MRI) analysis, contingent upon expert interpretation, grapples with challenges such as time-intensive processes and susceptibility to human error. This research presents a novel Convolutional Neural Network (CNN) architecture designed to enhance the accuracy and efficiency of brain tumor detection in MRI scans. The dataset used in the study comprises 7,023 brain MRI images from figshare, SARTAJ, and Br35H, categorized into glioma, meningioma, no tumor, and pituitary classes, with a CNN-based multi-task classification model employed for tumor detection, classification, and location identification. Our methodology focused on multi-task classification using a single CNN model for various brain MRI classification tasks, including tumor detection, classification based on grade and type, and tumor location identification. The proposed CNN model incorporates advanced feature extraction capabilities and deep learning optimization techniques, culminating in a groundbreaking paradigm shift in automated brain MRI analysis. With an exceptional tumor classification accuracy of 99%, our method surpasses current methodologies, demonstrating the remarkable potential of deep learning in medical applications. This study represents a significant advancement in the early detection and treatment planning of brain tumors, offering a more efficient and accurate alternative to traditional MRI analysis methods.

Cross-Modality Reference and Feature Mutual-Projection for 3D Brain MRI Image Super-Resolution.

High-resolution (HR) magnetic resonance imaging (MRI) can reveal rich anatomical structures for clinical diagnoses. However, due to hardware and signal-to-noise ratio limitations, MRI images are often collected with low resolution (LR) which is not conducive to diagnosing and analyzing clinical diseases. Recently, deep learning super-resolution (SR) methods have demonstrated great potential in enhancing the resolution of MRI images; however, most of them did not take the cross-modality and internal priors of MR seriously, which hinders the SR performance. In this paper, we propose a cross-modality reference and feature mutual-projection (CRFM) method to enhance the spatial resolution of brain MRI images. Specifically, we feed the gradients of HR MRI images from referenced imaging modality into the SR network to transform true clear textures to LR feature maps. Meanwhile, we design a plug-in feature mutual-projection (FMP) method to capture the cross-scale dependency and cross-modality similarity details of MRI images. Finally, we fuse all feature maps with parallel attentions to produce and refine the HR features adaptively. Extensive experiments on MRI images in the image domain and k-space show that our CRFM method outperforms existing state-of-the-art MRI SR methods.

The clinical utility of a circulating tumor cell based cerebrospinal fluid assay in the diagnosis and molecular analysis of leptomeningeal disease in patients with metastatic non-small cell lung cancer.

8644 Background: Leptomeningeal disease (LMD) is associated with significant morbidity and mortality for metastatic non-small cell lung cancer (mNSCLC). The standard of care for diagnosing LMD remains magnetic resonance imaging (MRI) and cerebrospinal fluid (CSF) cytology, but these have limited sensitivity. CSF-based liquid biopsy is an emerging strategy for dynamic assessments of CSF-derived tumor cells to monitor intracranial responses and CNS clonal evolution during treatment and at progression. However, data for the feasibility and utility of this approach is limited. We describe our clinical experience to evaluate the use of molecular analysis on CSF-derived circulating tumor cells (CTCs) for mNSCLC. Methods: Patients with mNSCLC who had a lumbar puncture for CSF collection as part of routine clinical care to rule out LMD were included in the study. CSF was evaluated for CTCs and cell-free DNA (cfDNA) with a commercially available assay (CNSide) and subsequent molecular profiling was performed to detect actionable mutations. Patients were analyzed for LMD with CSF cytology and MRI imaging. Molecular testing results from sequencing of tumor tissue and plasma cell tumor DNA were available for all patients. Other than copy number amplification detection for MET, the CSF assay allowed for analysis of MET and HER2 expression via fluorescent in situ hybridization (FISH). Results: The study included 22 patients with mNSCLC (77% female; median age 60 years; 95% adenocarcinoma). Sensitizing EGFR mutations were found in 78% of patients and 18% had an atypical EGFR mutation at initial diagnosis. LMD was diagnosed using the CSF CTC assay in thirteen of the 22 patients (59%) patients. CSF cytology was negative for LMD in five of 13 cases (38%) and furthermore, 2 of these cases (15%) had normal MRI brain imaging. There were sufficient CTCs to perform molecular profiling on seven of the 13 patients (54%). The concordance with tissue NGS was 100% and the driver mutation was identified in all 7 patients with the CSF CTC assay. MET expression and HER2 expression via FISH were noted in 11 patients (50%) and 4 patients (18%) respectively. Conclusions: In this single institution study, we detected a higher sensitivity to diagnose LMD with CNSide. This CSF assay identified 38% of LMD cases which were missed by standard CSF cytology and of note, 15% of the cases found with LMD were missed by both CSF cytology and MRI. In addition, CSF molecular testing using this assay demonstrated high concordance with tissue-based molecular testing. The study illustrates that multigene molecular analysis of CSF CTCs can be a useful tool for diagnosis, monitoring of response to therapy, as well as for the identification of therapeutic targets specific to the CSF especially for patients with oncogene-driven mNSCLC.