Magnetic Resonance Imaging Tumor Research Articles

Multi-resolution dual-encoder self-constrained brain tumor MR image segmentation algorithm

Multi-resolution dual-encoder self-constrained brain tumor MR image segmentation algorithm

Multi-scale feature fusion and additive attention guide brain tumor MR image segmentation

Multi-scale feature fusion and additive attention guide brain tumor MR image segmentation

Multiclass convolutional neural network based classification for the diagnosis of brain MRI images

Magnetic Resonance Imaging (MRI) modality is commonly used by radiologists in the medical field to diagnose an abnormality in the brain. The manual diagnosis process of grade identification is a challenging task that may lead to a False Negative, False Positive and it is also difficult to diagnose the early stages of abnormality. Manual misprediction may reduce the chance of survival in people with brain abnormalities. Hence, computer-aided diagnosis in additional supports the radiologist to identify the abnormality accurately even in the early stages of a brain tumor. In our proposed system, we have developed a Multi Class Convolutional Neural Network model (MCCNN) to identify the tumor in Brain MRI. In this work, BRATS 2015 and Figshare Data are used. After pre-processing, the feature vector is constructed from MRI using the convolution and pooling layers of CNN. Finally, the softmax layer of CNN identifies the tumor. For analyzing the performance of the proposed MCCNN model, two experiments such as Experiment I and Experiment II are conducted. The designed MCCNN provides less complexity when compared to pre-trained models and other CNN-based networks in literature. It decreases loss value, False Negative Rate, False Positive Rate and increases the accuracy of classification. The proposed MCCNN-based model attained a noteworthy performance with an accuracy of 99% in Experiment I and an accuracy of 96% in Experiment II.

Brain Tumor Classification for MR Images Using Transfer Learning and EfficientNetB3

Brain tumors are one of the diseases that take many lives in the world, moreover, brain tumors have various types. In the medical world, it has an technology called Magnetic Resonance Imaging (MRI) which functions to see the inside of the human body using a magnetic field. CNN is designed to determine features adaptively using backpropagation by applying layers such as convolutional layers, and pooling layers. This study aims to optimize and increase the accuracy of the classification of brain tumor MRI images using the Convolutional Neural Network (CNN) EfficientNet model. The proposed system consists of two main steps. First, preprocessing images using various methods then classifying images that have been preprocessed using CNN. This study used 3064 images containing three types of brain tumors (gliomata, meningiomas, and pituitary). This study resulted in an accuracy of 98.00%, a precision of 96.00%, and an average recall of 97.00% using the model that the researcher applied. 

Hypoxia-Responsive Aggregation of Iron Oxide Nanoparticles for T1-to-T2 Switchable Magnetic Resonance Imaging of Tumors

Magnetic resonance imaging (MRI), one of the most general imaging techniques in the clinic, is widely employed for tumor diagnosis with the assistance of contrast agents (CAs). However, currently available CAs in the clinic suffer from shortcomings of a short blood circulation time, lack of specificity to pathological location, insufficient imaging window, and potential biotoxicity of the gadolinium ion. To overcome these problems, on the basis of extremely small iron oxide nanoparticles (ESIONPs), a hypoxia-responsive T1-to-T2 switchable MRI contrast agent was designed and constructed. In detail, polyacrylic acid-modified ESIONPs (ESIONPs-PAA) were first synthesized, and then, nitroimidazole derivatives and cysteine molecules were sequentially coupled on the surface of ESIONPs-PAA to obtain a hypoxia-responsive MRI contrast agent (HR-ESIONPs). Since 2-nitroimidazole and cysteine molecules on the surface of HR-ESIONPs can form an irreversible bond induced by the nitroreductase in the hypoxic tumor microenvironment, monodispersed HR-ESIONPs could transform into the nanocluster form, which can specially provide T2 contrast enhancement. In vitro experiments show that monodispersed HR-ESIONPs could effectively aggregate in the mimical hypoxic environment of tumors and switch from a T1 to T2 contrast agent. Moreover, in vivo experiments further confirm that HR-ESIONPs could effectively accumulate and realize the MRI signal switch from T1 to T2 contrast enhancement in tumors originating from the aggregation of HR-ESIONPs induced by the hypoxic environment, which promotes the diagnostic precision of tumors. Additionally, with the confirmation of favorable biosafety on cells and mice, hypoxia-responsive HR-ESIONPs are potential platforms to improve the diagnostic accuracy and sensitivity of MRI for tumors.

Nanoprobe Based on Biominerals in Protein Corona for Dual-Modality MR Imaging and Therapy of Tumors.

Various functional nanomaterials have been fabricated as diagnostic and therapeutic nanomedicines; however, the nanoparticles closely interact with proteins when immersed in biological fluids, forming a "protein corona" that critically alters the biological identity of nanomedicine. Here, we developed a robust strategy to construct theranostic nanoprobes based on protein-corona-coated Fe3O4 nanoparticles and biomineralization in the corona. Water-soluble carboxylic Fe3O4 nanoparticles were prepared by treating oleate-capped Fe3O4 nanoparticles with Lemieux-von Rudloff reagent. Bovine serum albumin (BSA) was used as a model protein to form a corona on the surface of Fe3O4 nanoparticles, endowing the Fe3O4 nanoparticles with biocompatibility and nonimmunogenicity. The protein corona also provides a template for biomimetic mineralization of Fe3+ with tannic acid (TA) to construct Fe3O4@BSA-TAFeIII nanoprobes. The TA-Fe(III) biominerals can not only act as photothermal therapy agents but also interact with unsaturated transferrin in plasma to form a "hybrid" corona, enabling the nanoprobes to target tumor cells through the mediation of transferrin receptors, which commonly overexpress on tumor cell membranes. Once taken in by tumor cells, the protonation of phenol hydroxyl groups in acidic lysosomes would lead to the release of Fe3+, inducing tumor cell death through a ferroptosis/apoptosis hybrid pathway. In addition, the released Fe3+ can boost the T1-weighted MR imaging performance, and the Fe3O4 nanoparticles serve as T2-weighted MR imaging contrast agents. It is thus believed that the current nanoprobes can realize the enhanced dual-modality MR imaging and combined therapy of tumors through controlling the protein corona and biomineralization.

Tumor microenvironment-responsive histidine modified-hyaluronic acid-based MnO2 as in vivo MRI contrast agent

Tumor microenvironment (TME)-responsive manganese dioxide (MnO2) nanoparticles as a good T1 contrast agent could reduce unwanted toxicity and improve the accuracy of cancer detection. Despite these distinct advantages of MnO2-based nanoparticles, their synthesis involves multi-step processes with relatively long synthesis times. In this study, we synthesized histidine-modified hyaluronic acid (HA-His), and the prepared HA-His conjugates quickly reduce permanganate to MnO2, leading to facile production of HA-His/MnO2 nanoparticles with good water-dispersibility and stability under biological conditions. The synthesized HA-His/MnO2 nanoparticles readily responded to the TME (low pH, high H2O2, and high glutathione), and they were internalized into SCC7 cells with high CD44 expression. Moreover, the systemically administered HA-His/MnO2 nanoparticles with biocompatibility were specifically accumulated in tumor tissues, thereby efficiently enhancing T1 contrast in MRI. Therefore, the HA-His/MnO2 nanoparticles synthesized herein can be used as a promising T1 contrast agent for tumor MR imaging.

Ultrasmall Fe3O4 and Gd2O3 hybrid nanoparticles for T1-weighted MR imaging of cancer

AbstractGadolinium-based contrast agents (GdCAs) have been the most frequently used T1-weighted magnetic resonance imaging (MRI) contrast agents for decades. However, they have serious disadvantages such as low longitudinal relaxivity value (r1) and high dose associated-nephrotoxicity that restrict their wide applications. These emphasize the need for an ideal stable and biocompatible T1-weighted CA with high contrast enhancement performance. Here, we propose a wet-chemical synthesis technique to form a nanocomposite consisting of ultrasmall iron oxide nanoparticles (US-IO) and Gd2O3 hybrid nanoparticles stabilized with dextran (FG-HNPs) for T1-weighted MR imaging. Relaxometry study showed that FG-HNPs have a high r1 value (42.28 mM−1S−1) and low relaxivity ratio (r2/r1: 1.416) at 3.0T. In vivo MRI contrast enhancement factor (ΔSNR) for FG-HNPs (257.025 ± 17.4%) was found to be 1.99-fold higher than US-IO (129.102 ± 15%) and 3.35-fold higher than Dotarem (76.71 ± 14.2%) as routinely used T1-weighted CA. The cytotoxicity assay and histological examination confirmed the biocompatibility of FG-HNPs. The biodistribution study, transmission electron microscopy (TEM) and Prussian blue (PB) staining of tumor tissue proved the effective tumor localization of FG-HNPs. Therefore, FG-HNPs can be suggested as a promising CA for T1-weighted MRI of tumors by virtue of their remarkable relaxivities and high biocompatibility.

Deep Learning-Based Segmentation and Classification Techniques for Brain Tumor MRI: A Review

Early detection of brain tumors is critical for enhancing treatment options and extending patient survival. Magnetic resonance imaging (MRI) scanning gives more detailed information, such as greater contrast and clarity than any other scanning method. Manually dividing brain tumors from many MRI images collected in clinical practice for cancer diagnosis is a tough and time-consuming task. Tumors and MRI scans of the brain can be discovered using algorithms and machine learning technologies, making the process easier for doctors because MRI images can appear healthy when the person may have a tumor or be malignant. Recently, deep learning techniques based on deep convolutional neural networks have been used to analyze medical images with favorable results. It can help save lives faster and rectify some medical errors. In this study, we look at the most up-to-date methodologies for medical image analytics that use convolutional neural networks on MRI images. There are several approaches to diagnosing and classifying brain cancers. Inside the brain, irregular cells grow so that a brain tumor appears. The size of the tumor and the part of the brain affected impact the symptoms.

Clinical Decision Support Framework for Segmentation and Classification of Brain Tumor MRIs Using a U-Net and DCNN Cascaded Learning Algorithm.

Brain tumors (BTs) are an uncommon but fatal kind of cancer. Therefore, the development of computer-aided diagnosis (CAD) systems for classifying brain tumors in magnetic resonance imaging (MRI) has been the subject of many research papers so far. However, research in this sector is still in its early stage. The ultimate goal of this research is to develop a lightweight effective implementation of the U-Net deep network for use in performing exact real-time segmentation. Moreover, a simplified deep convolutional neural network (DCNN) architecture for the BT classification is presented for automatic feature extraction and classification of the segmented regions of interest (ROIs). Five convolutional layers, rectified linear unit, normalization, and max-pooling layers make up the DCNN's proposed simplified architecture. The introduced method was verified on multimodal brain tumor segmentation (BRATS 2015) datasets. Our experimental results on BRATS 2015 acquired Dice similarity coefficient (DSC) scores, sensitivity, and classification accuracy of 88.8%, 89.4%, and 88.6% for high-grade gliomas. When it comes to segmenting BRATS 2015 BT images, the performance of our proposed CAD framework is on par with existing state-of-the-art methods. However, the accuracy achieved in this study for the classification of BT images has improved upon the accuracy reported in prior studies. Image classification accuracy for BRATS 2015 BT has been improved from 88% to 88.6%.

A dynamic nomogram combining tumor stage and magnetic resonance imaging features to predict the response to induction chemotherapy in locally advanced nasopharyngeal carcinoma.

To establish an effective dynamic nomogram combining magnetic resonance imaging (MRI) findings of primary tumor and regional lymph nodes with tumor stage for the pretreatment prediction of induction chemotherapy (IC) response in locoregionally advanced nasopharyngeal carcinoma (LANPC). A total of 498 LANPC patients (372 in the training and 126 in the validation cohort) with MRI information were enrolled. All patients were classified as "favorable responders" and "unfavorable responders" according to tumor response to IC. A nomogram for IC response was built based on the results of the logistic regression model. Also, the Cox regression analysis was used to identify the independent prognostic factors of disease-free survival (DFS). After two cycles of IC, 340 patients were classified as "favorableresponders" and 158 patients as "unfavorableresponders." Calibration curves revealed satisfactory agreement between the predicted and the observed probabilities. The nomogram achieved an AUC of 0.855 (95% CI, 0.781-0.930) for predicting IC response, which outperformed TNM staging (AUC, 0.661; 95% CI 0.565-0.758) and the MRI feature-based model alone (AUC, 0.744; 95% CI 0.650-0.839) in the validation cohort. The nomogram was used to categorize patients into high- and low-response groups. An online dynamic model was built ( https://nomogram-for-icresponse-prediction.shinyapps.io/DynNomapp/ ) to facilitate the application of the nomogram. In the Cox multivariate analysis, clinical stage, tumor necrosis, EBV DNA levels, and cervical lymph node numbers were independently associated with DFS. The comprehensive nomogram incorporating MRI features and tumor stage could assist physicians in predicting IC response and formulating personalized treatment strategies for LANPC patients. • The nomogram can predict IC response in endemic LANPC. • The nomogram combining tumor stage with MRI-based tumor features showed very good predictive performance. • The nomogram was transformed into a web-based dynamic model to optimize clinical application.

Supratentorial Meningioma Consistency Prediction Utilizing Tumor to Cerebellar Peduncle Intensity on T1 and T2-Weighted and Fluid Attenuated Inversion Recovery Magnetic Resonance Imaging Sequences

Predicting meningioma consistency with preoperative imaging is critical for surgery planning. Preoperative T1 and T2-weighted and fluid attenuated inversion recovery magnetic resonance imaging (MRI) findings of supratentorial meningioma tumors were studied and compared with intraoperative supratentorial meningioma tumor consistency based on the Cavitron ultrasound surgical aspirator (CUSA) and ZADA grading scales in this cohort to predict the tumor consistency before surgery. MRI from 78 consecutive patients who underwent supratentorial meningioma tumor resection between 2018 and 2021 were evaluated preoperatively. An intraoperative tumor consistency grade was applied to these lesions prospectively by the operating surgeon based on CUSA and ZADA grading scales. Tumor/cerebellar peduncle T2-weighted intensity, tumor/cerebellar peduncle T1-weighted intensity (TCT1I), and tumor/cerebellar peduncle fluid attenuated inversion recovery intensity (TCFI) ratios were calculated. Tumor consistency grades and MRI intensity ratios were correlated using one-way ANOVA. Of the 78 patients, 52 (66.7%) were female and 26 (33.3%) were male. Tumor volume correlated with tumor consistency grades on both CUSA (P= 0.005) and ZADA (P= 0.024) grading scales. Also patients age correlated with tumor consistency according to ZADA grading scale (P= 0.024). TCT1I (P= 0.009) and TCFI (P < 0.005) ratios correlated significantly with tumor consistency grade according to CUSA. Similarly, TCT1I (P= 0.0032) and TCFI (P= 0.001) ratios was significantly associated with tumor consistency according to ZADA grading scales. Our findings suggest that higher tumor/cerebellar peduncle T2-weighted intensity and TCFI ratios correlate with softer tumors, while higher TCT1I ratios reveal firmer tumors. These data can assist the surgeon predict the supratentorial meningioma consistency before surgery.

Liver Function Preservation and Outcome after SBRT with Functional Treatment Planning in Patients with HCC and Advanced Hepatic Cirrhosis: From Nuclear Medicine to MRI-Based Functional Treatment Planning Platform

<h3>Purpose/Objective(s)</h3> To evaluate treatment outcomes of ablative SBRT with SPECT/CT and MRI-guided functional treatment planning for localized HCC in patients with Child-Pugh B (CP-B) cirrhosis. <h3>Materials/Methods</h3> Liver SPECT with ^99mTc-sulfur colloid was co-registered to 3D-CT for identification and avoidance of functional hepatic parenchyma during SBRT in patients with advanced cirrhosis. Likewise, MRI scanner of the Unity® MR-Linac with super-paramagnetic iron oxide nanoparticle (SPION) agent Feraheme® as an alternative contrast was used for treatment planning with visualization of tumor and functional liver parenchyma during SBRT. Functional liver volumes (FLV) obtained from CT-SPECT or SPION-MRI were compared to anatomical liver volumes, as were dosimetric parameters when radiation dose constraints were adapted exclusively to FLV. Hepatic function, toxicity, and radiographic response were documented every 3-4 months following SBRT. <h3>Results</h3> With a median follow-up of 25 months, 44 patients (59 lesions treated) with CP-B cirrhosis received ablative SBRT to a median dose of 48 Gy (3–5 fractions). An automated contouring algorithm was generated for delineation and guided avoidance of best functional hepatic parenchyma on SPION-MRI. Prolonged SPION-contrast retention within hepatic parenchyma allowed per fraction treatment plan adaptation with SPION-enhanced MR imaging of tumor throughout the treatment course. FLV loss (509 cc or 41.3%, p<0.001) was observed in all patients, while the functional and anatomical liver volumes matched well in a control group of non-cirrhotic/non-HCC patients. Mean dose to residual FLV was maintained well below the liver threshold tolerance to radiation in all patients. Thirty-six transplant eligible patients received SBRT as a bridging treatment and 26 patients (60% of the cohort) were listed for liver transplant after completion of SBRT. Twenty-two patients successfully completed orthotopic liver transplant (OLT) with a median time to transplant of 6.1 months. Eight of 10 patients with intrahepatic progression received additional SBRT during follow-up and four of them completed transplant. With 95% in-field local control rate, overall, 2-year survival was 63% (90% after the OLT) with no incidence of RILD or CP class migration observed at 6+ months post SBRT. <h3>Conclusion</h3> SBRT with functional treatment planning on SPEC/CT allows identification and guided avoidance of residual functionally active hepatic parenchyma in patients with advanced hepatic cirrhosis. SPION-enhanced MRI provides consistent visibility of liver tumors and functional hepatic parenchyma on the day of treatment planning and throughout the online MR-guided SBRT. We report high local control and low toxicity leading to satisfactory pre- and post-liver transplant outcomes. Prospective clinical trial investigating MRI-SPION adaptive planning on the 1.5T Elekta Unity® MR-Linac is ongoing in our institution.

Chemotherapy Mediated by Biomimetic Polymeric Nanoparticles Potentiates Enhanced Tumor Immunotherapy via Amplification of Endoplasmic Reticulum Stress and Mitochondrial Dysfunction.

Construction of multifunctional nanoplatforms to elevate chemotherapeutic efficacy and induce long-term antitumor immunity still remains to be an extreme challenge. Herein, the design of an advanced redox-responsive nanomedicine formulation based on phosphorus dendrimer-copper(II) complexes (1G3 -Cu)- and toyocamycin (Toy)-loaded polymeric nanoparticles (GCT NPs) coated with cancer cell membranes (CM) are reported. The designed GCT@CM NPs with a size of 210 nm are stable under physiological conditions but are rapidly dissociated in the reductive tumor microenvironment to deplete glutathione and release drugs. The co-loading of 1G3 -Cu and Toy within the NPs causes significant tumor cell apoptosis and immunogenic cell death through 1G3 -Cu-induced mitochondrial dysfunction and Toy-mediated amplification of endoplasmic reticulum stress, respectively, thus effectively suppressing tumor growth, promoting dendritic cell maturation, and increasing tumor-infiltrating cytotoxic T lymphocytes. Likewise, the coated CM and the loaded 1G3 -Cu render the GCT@CM NPs with homotypic targeting and T1 -weighted magnetic resonance imaging of tumors, respectively. With the assistance of programmed cell death ligand 1 antibody, the GCT@CM NP-mediated chemotherapy can significantly potentiate tumor immunotherapy for effective inhibition of tumor recurrence and metastasis. The developed GCT@CM NPs hold a great potential for chemotherapy-potentiated immunotherapy of different tumor types through different mechanisms or synergies.

MBANet: A 3D convolutional neural network with multi-branch attention for brain tumor segmentation from MRI images

More than half of brain tumors are malignant tumors, so there is a need for fast and accurate segmentation of tumor regions in brain Magnetic Resonance Imaging (MRI) images. Traditional 2D brain tumor segmentation methods seriously ignore the spatial context features of brain tumor MRI images, so how to achieve accurate segmentation of brain tumor regions with multiple modalities is the main problem. The paper proposes a 3D convolutional neural network with 3D multi-branch attention - MBANet. First, the optimized shuffle unit is used to form the basic unit (BU) module of MBANet. In the BU module, the group convolution is used to perform convolution operation after the input channel is split, and the channel shuffle is used to scramble the convolutional channels after fusion. Then, MBANet uses a novel multi-branch 3D Shuffle Attention (SA) module as the attention layer in the encoder. The 3D SA module groups along the channel dimension and divides the feature maps into small features. For each small feature, the 3D SA module builds both channel attention and spatial attention while adopting the BU module. In addition, in order to recover the resolution of the upsampling semantic features better, a 3D SA module is also used in the skip connection of MBANet. Experiments on the BraTS 2018 and BraTS 2019 show that the dice of ET, WT and TC reach 80.18%, 89.80%, 85.47% and 78.21%, 89.79%, 83.04%, respectively. The excellent segmentation performance shows that MBANet is significantly improved compared with other state-of-the-art methods.

Multifunctional PVCL nanogels enable magnetic resonance imaging and immunostimulated radiotherapy of orthotopic glioblastoma

Despite radiotherapy (RT) is considered as an indispensable treatment that improves clinical symptoms and survival rates of glioblastoma multiforme (GBM), the overall survival time of GBM based on multimodal treatment synergic with RT is still unsatisfying. Herein, we designed an immunostimulated nanoplatform consisting of MnO2 and cytosine-phosphorothioate-guanine oligodeoxynucleotides (CpG ODNs) within the poly(N-vinylcaprolactam) nanogels (PVCL-MnO2-CpG NGs) for magnetic resonance imaging (MRI)-guided immuno-activated RT against GBM. In this system, I) CpG ODNs can be protected by PVCL NGs from degradation, which can increase the cellular uptake by immune cells and promote the activation of Toll-like receptor 9 (TLR9) pathway, thus enhance the immune responses to assist the therapeutic efficacy of tumour RT. II) PVCL-MnO2-CpG NGs are able to be dissociated to release Mn2+ in the tumour microenvironment, resulting in excellent T1-weighted tumour MRI along with the activation of cyclic GMP–AMP synthase/stimulator of interferon genes (cGAS/STING) pathway and elevated secretion of type-I interferons (IFN-I) in the downstream. III) MnO2 can also act as nanoenzyme to relieve tumour radioresistance through increasing the reactive oxygen species (ROS) level. Compared with RT alone, combinational immunoradiotherapy remarkably promotes the proportion of cytotoxic T cells (CD3+CD8+) and exhibits systemic therapeutic efficacy. Thus, MnO2 and CpG ODNs-loaded PVCL NGs may serve as a potential theranostic platform for MRI-guided immunostimulated RT of GBM.

Diagnosing pancreatic neuroendocrine tumors in patients with multiple endocrine neoplasia type 1 in daily practice.

BackgroundIn multiple endocrine neoplasia type 1 (MEN1), pancreatic neuroendocrine tumors (PanNETs) have a high prevalence and represent the main cause of death. This study aimed to assess the diagnostic accuracy of the currently used conventional pancreatic imaging techniques and the added value of fine needle aspirations (FNAs).MethodsPatients who had at least one imaging study were included from the population-based MEN1 database of the DutchMEN Study Group from 1990 to 2017. Magnetic resonance imaging (MRI), computed tomography (CT), endoscopic ultrasonography (EUS), FNA, and surgical resection specimens were obtained. The first MRI, CT, or EUS was considered as the index test. For a comparison of the diagnostic accuracy of MRI versus CT, patients with their index test taken between 2010 and 2017 were included. The reference standard consisted of surgical histopathology or radiological follow-up.ResultsA total of 413 patients (92.8% of the database) underwent 3,477 imaging studies. The number of imaging studies per patient increased, and a preference for MRI was observed in the last decade. Overall diagnostic accuracy was good with a positive (PPV) and negative predictive value (NPV) of 88.9% (95% confidence interval, 76.0–95.6) and 92.8% (89.4–95.1), respectively, for PanNET in the pancreatic head and 92.0% (85.3–96.0) and 85.3% (80.5–89.1), respectively, in the body/tail. For MRI, PPV and NPV for pancreatic head tumors were 100% (76.1–100) and 87.1% (76.3–93.6) and for CT, 60.0% (22.9–88.4) and 70.4% (51.3–84.3), respectively. For body/tail tumors, PPV and NPV were 91.3% (72.0–98.8) and 87.0% (75.3–93.9), respectively, for MRI and 100% (74.9–100) and 77.8% (54.3–91.5), respectively, for CT. Pathology confirmed a PanNET in 106 out of 110 (96.4%) resection specimens. FNA was performed on 34 lesions in 33 patients and was considered PanNET in 24 [all confirmed PanNET by histology (10) or follow-up (14)], normal/cyst/unrepresentative in 6 (all confirmed PanNET by follow-up), and adenocarcinoma in 4 (2 confirmed and 2 PanNET). Three patients, all older than 60 years, had a final diagnosis of pancreatic adenocarcinoma.ConclusionAs the accuracy for diagnosing MEN1-related PanNET of MRI was higher than that of CT, MRI should be the preferred (non-invasive) imaging modality for PanNET screening/surveillance. The high diagnostic accuracy of pancreatic imaging and the sporadic occurrence of pancreatic adenocarcinoma question the need for routine (EUS-guided) FNA.

A Robust End-to-End Deep Learning-Based Approach for Effective and Reliable BTD Using MR Images.

Detection of a brain tumor in the early stages is critical for clinical practice and survival rate. Brain tumors arise in multiple shapes, sizes, and features with various treatment options. Tumor detection manually is challenging, time-consuming, and prone to error. Magnetic resonance imaging (MRI) scans are mostly used for tumor detection due to their non-invasive properties and also avoid painful biopsy. MRI scanning of one patient’s brain generates many 3D images from multiple directions, making the manual detection of tumors very difficult, error-prone, and time-consuming. Therefore, there is a considerable need for autonomous diagnostics tools to detect brain tumors accurately. In this research, we have presented a novel TumorResnet deep learning (DL) model for brain detection, i.e., binary classification. The TumorResNet model employs 20 convolution layers with a leaky ReLU (LReLU) activation function for feature map activation to compute the most distinctive deep features. Finally, three fully connected classification layers are used to classify brain tumors MRI into normal and tumorous. The performance of the proposed TumorResNet architecture is evaluated on a standard Kaggle brain tumor MRI dataset for brain tumor detection (BTD), which contains brain tumor and normal MR images. The proposed model achieved a good accuracy of 99.33% for BTD. These experimental results, including the cross-dataset setting, validate the superiority of the TumorResNet model over the contemporary frameworks. This study offers an automated BTD method that aids in the early diagnosis of brain cancers. This procedure has a substantial impact on improving treatment options and patient survival.

MR Imaging and Clinical Characteristics of Diffuse Glioneuronal Tumor with Oligodendroglioma-like Features and Nuclear Clusters.

Diffuse glioneuronal tumor with oligodendroglioma-like features and nuclear clusters (DGONC) is a new, molecularly defined glioneuronal CNS tumor type. The objective of the present study was to describe MR imaging and clinical characteristics of patients with DGONC. Preoperative MR images of 9 patients with DGONC (median age at diagnosis, 9.9 years; range, 4.2-21.8 years) were reviewed. All tumors were located superficially in the frontal/temporal lobes and sharply delineated, displaying little mass effect. Near the circle of Willis, the tumors encompassed the arteries. All except one demonstrated characteristics of low-to-intermediate aggressiveness with high-to-intermediate T2WI and ADC signals and bone remodeling. Most tumors (n = 7) showed a homogeneous ground-glass aspect on T2-weighted and FLAIR images. On the basis of the original histopathologic diagnosis, 6 patients received postsurgical chemo-/radiotherapy, 2 were irradiated after surgery, and 1 patient underwent tumor resection only. At a median follow-up of 61 months (range, 10-154 months), 6 patients were alive in a first complete remission and 2 with stable disease 10 and 21 months after diagnosis. The only patient with progressive disease was lost to follow-up. Five-year overall and event-free survival was 100% and 86±13%, respectively. This case series presents radiomorphologic characteristics highly predictive of DGONC that contrast with the typical aspects of the original histopathologic diagnoses. This presentation underlines the definition of DGONC as a separate entity, from a clinical perspective. Complete resection may be favorable for long-term disease control in patients with DGONC. The efficacy of nonsurgical treatment modalities should be evaluated in larger series.

Edge U-Net: Brain tumor segmentation using MRI based on deep U-Net model with boundary information

Blood clots in the brain are frequently caused by brain tumors. Early detection of these clots has the potential to significantly lower morbidity and mortality in cases of brain cancer. It is thus indispensable for a proper brain tumor diagnosis and treatment that tumor tissue magnetic resonance images (MRI) be accurately segmented. Several deep learning approaches to the segmentation of brain tumor MRIs have been proposed, each having been designed to properly map out ‘boundaries’ and thus achieve highly accurate segmentation. This study introduces a deep convolution neural network (DCNN), named the Edge U-Net model, built as an encoder-decoder structure inspired by the U-Net architecture. The Edge U-Net model can more precisely localise tumors by merging boundary-related MRI data with the main data from brain MRIs. In the decoder phase, boundary-related information from original MRIs of different scales is integrated with the appropriate adjacent contextual information. A novel loss function was added to this segmentation model to improve performance. This loss function is enhanced with boundary information, allowing the learning process to produce more precise results. In the conducted experiments, a public dataset with 3064 T1-Weighted Contrast Enhancement (T1-CE) images of three well-known brain tumor types were used. The experiment demonstrated that the proposed framework achieved satisfactory Dice score values compared with state-of-art models, with highly accurate differentiation of brain tissues. It achieved Dice scores of 88.8 % for meningioma, 91.76 % for glioma, and 87.28 % for pituitary tumors. Computations of other performance metrics such as the Jaccard index, sensitivity, and specificity were also conducted. According to the results, the Edge U-Net model is a potential diagnostic tool that can be used to help radiologists more precisely segment brain tumor tissue images.