Attention 3D U-Net with Multiple Skip Connections for Segmentation of Brain Tumor Images

Among researchers using traditional and new machine learning and deep learning techniques, 2D medical image segmentation models are popular. Additionally, 3D volumetric data recently became more accessible, as a result of the high number of studies conducted in recent years regarding the creation of 3D volumes. Using these 3D data, researchers have begun conducting research on creating 3D segmentation models, such as brain tumor segmentation and classification. Since a higher number of crucial features can be extracted using 3D data than 2D data, 3D brain tumor detection models have increased in popularity among researchers. Until now, various significant research works have focused on the 3D version of the U-Net and other popular models, such as 3D U-Net and V-Net, while doing superior research works. In this study, we used 3D brain image data and created a new architecture based on a 3D U-Net model that uses multiple skip connections with cost-efficient pretrained 3D MobileNetV2 blocks and attention modules. These pretrained MobileNetV2 blocks assist our architecture by providing smaller parameters to maintain operable model size in terms of our computational capability and help the model to converge faster. We added additional skip connections between the encoder and decoder blocks to ease the exchange of extracted features between the two blocks, which resulted in the maximum use of the features. We also used attention modules to filter out irrelevant features coming through the skip connections and, thus, preserved more computational power while achieving improved accuracy.

Fuzzy System Based Medical Image Processing for Brain Disease Prediction

The present work aims to explore the performance of fuzzy system-based medical image processing for predicting the brain disease. The imaging mechanism of NMR (Nuclear Magnetic Resonance) and the complexity of human brain tissues cause the brain MRI (Magnetic Resonance Imaging) images to present varying degrees of noise, weak boundaries, and artifacts. Hence, improvements are made over the fuzzy clustering algorithm. A brain image processing and brain disease diagnosis prediction model is designed based on improved fuzzy clustering and HPU-Net (Hybrid Pyramid U-Net Model for Brain Tumor Segmentation) to ensure the model safety performance. Brain MRI images collected from a Hospital, are employed in simulation experiments to validate the performance of the proposed algorithm. Moreover, CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), FCM (Fuzzy C-Means), LDCFCM (Local Density Clustering Fuzzy C-Means), and AFCM (Adaptive Fuzzy C-Means) are included in simulation experiments for performance comparison. Results demonstrate that the proposed algorithm has more nodes, lower energy consumption, and more stable changes than other models under the same conditions. Regarding the overall network performance, the proposed algorithm can complete the data transmission tasks the fastest, basically maintaining at about 4.5 s on average, which performs remarkably better than other models. A further prediction performance analysis reveals that the proposed algorithm provides the highest prediction accuracy for the Whole Tumor under DSC (Dice Similarity Coefficient), reaching 0.936. Besides, its Jaccard coefficient is 0.845, proving its superior segmentation accuracy over other models. In a word, the proposed algorithm can provide higher accuracy, a more apparent denoising effect, and the best segmentation and recognition effect than other models while ensuring energy consumption. The results can provide an experimental basis for the feature recognition and predictive diagnosis of brain images.

Semi-Supervised Support Vector Machine for Digital Twins Based Brain Image Fusion

The purpose is to explore the feature recognition, diagnosis, and forecasting performances of Semi-Supervised Support Vector Machines (S3VMs) for brain image fusion Digital Twins (DTs). Both unlabeled and labeled data are used regarding many unlabeled data in brain images, and semi supervised support vector machine (SVM) is proposed. Meantime, the AlexNet model is improved, and the brain images in real space are mapped to virtual space by using digital twins. Moreover, a diagnosis and prediction model of brain image fusion digital twins based on semi supervised SVM and improved AlexNet is constructed. Magnetic Resonance Imaging (MRI) data from the Brain Tumor Department of a Hospital are collected to test the performance of the constructed model through simulation experiments. Some state-of-art models are included for performance comparison: Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), AlexNet, and Multi-Layer Perceptron (MLP). Results demonstrate that the proposed model can provide a feature recognition and extraction accuracy of 92.52%, at least an improvement of 2.76% compared to other models. Its training lasts for about 100 s, and the test takes about 0.68 s. The Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) of the proposed model are 4.91 and 5.59%, respectively. Regarding the assessment indicators of brain image segmentation and fusion, the proposed model can provide a 79.55% Jaccard coefficient, a 90.43% Positive Predictive Value (PPV), a 73.09% Sensitivity, and a 75.58% Dice Similarity Coefficient (DSC), remarkably better than other models. Acceleration efficiency analysis suggests that the improved AlexNet model is suitable for processing massive brain image data with a higher speedup indicator. To sum up, the constructed model can provide high accuracy, good acceleration efficiency, and excellent segmentation and recognition performances while ensuring low errors, which can provide an experimental basis for brain image feature recognition and digital diagnosis.

Abnormal Large-Scale Network Activation Present in Bipolar Mania and Bipolar Depression Under Resting State

Introduction: Previous studies have primarily focused on the neuropathological mechanisms of the emotional circuit present in bipolar mania and bipolar depression. Recent studies applying resting-state functional magnetic resonance imaging (fMRI) have raise the possibility of examining brain-wide networks abnormality between the two oppositional emotion states, thus this study aimed to characterize the different functional architecture represented in mania and depression by employing group-independent component analysis (gICA). Materials and Methods: Forty-one bipolar depressive patients, 20 bipolar manic patients, and 40 healthy controls (HCs) were recruited and received resting-state fMRI scans. Group-independent component analysis was applied to the brain network functional connectivity analysis. Then, we calculated the correlation between the value of between-group differences and clinical variables. Results: Group-independent component analysis identified 15 components in all subjects, and ANOVA showed that functional connectivity (FC) differed significantly in the default mode network, central executive network, and frontoparietal network across the three groups. Further post-hoc t-tests showed a gradient descent of activity-depression > HC > mania-in all three networks, with the differences between depression and HCs, as well as between depression and mania, surviving after family wise error (FWE) correction. Moreover, central executive network and frontoparietal network activities were positively correlated with Hamilton depression rating scale (HAMD) scores and negatively correlated with Young manic rating scale (YMRS) scores. Conclusions: Three brain networks heighten activity in depression, but not mania; and the discrepancy regions mainly located in prefrontal, which may imply that the differences in cognition and emotion between the two states is associated with top-down regulation in task-independent networks.

Decreased regional brain activity in response to sleep-related sounds after cognitive behavioral therapy for psychophysiological insomnia

AIM

Patients with psychophysiological insomnia (PI) experience hyperarousal, especially as a reaction to sound stimuli. In the current study, we explored brain activity changes in response to sleep-related sounds (SS) in patients with insomnia after cognitive behavioral therapy for insomnia (CBT-I).

METHODS

In 14 drug-free PI patients, regional brain activity in response to SS, and to white noise sound (NS) as neutral stimuli, was investigated before and after individual CBT-I using functional magnetic resonance imaging. Blood oxygen level-dependent (BOLD) signals to SS and NS were compared before and after CBT-I. In addition, the association between clinical improvement after CBT-I and changes in brain activity in response to SS and NS was analyzed.

RESULTS

Compared with baseline, regional brain activity in response to SS after CBT-I decreased in the left middle temporal and left middle occipital gyrus. In regression analysis, a reduction in the Dysfunctional Beliefs and Attitudes about Sleep (DBAS) Scale score after CBT-I was associated with decrease in brain activity in response to SS in both thalami. However, brain activity in response to NS showed no BOLD signal changes and no association with DBAS change.

CONCLUSION

Cortical hyperactivity, which may cause hyperarousal in PI, was found to decrease after CBT-I. CBT-I targeting changes in beliefs and attitudes about sleep may induce its therapeutic effects by reducing thalamic brain activity in response to sleep-related stimuli.

[Numerical and Visual Evaluations of Compressed Sensing MRI Using 2D Cartesian Sampling]

This paper describes numerical and visual evaluations of compressed sensing MRI (CS-MRI) using 2D Cartesian sampling by numerical simulation. The BrainWeb MRI Data Base was used for test images. Three brain anatomical ROIs (white matter, gray matter, cerebrospinal fluid) of a T1-weighted image (T1WI), a T2-weighted image (T2WI) and a proton density-weighted image (PDWI) were used for the numerical evaluation. Sampling ratio was 50%. Reconstruction was performed by minimizing the L1 norm of a transformed image using wavelet transform and total variation, subject to data fidelity constraints. The conjugate gradient method was used in the minimization of the object function. In the absence of noise, the root mean square error (RMSE) of T1WI was in the range of 2.99 to 3.57; that of the anatomical region of interests (ROIs) was in the range of 1.77 to 8.53; those of T2WI were 4.72 to 5.65 and 3.28 to 5.54; and those of PDWI were 1.91 to 2.36 and 1.32 to 2.09. Visual evaluation was performed by three radiologists on the basis of three categories: artifact, anatomical structure, tissue contrast. CS image quality was nearly equal to that of the original image, although a few artifacts were visible. If the noise level was assumed to be 30 dB or less, T1-CS image and PD-CS images were not significantly degraded compared to noise-free images.

[Numerical and Visual Evaluation of Compressed Sensing MRI Using 3D Cartesian Sampling]

We performed numerical and visual evaluation of compressed sensing MRI (CS-MRI) using 3D Cartesian sampling by numerical simulation. Three brain anatomical ROIs (white matter, gray matter, cerebrospinal fluid) of a T1-weighted image (T1WI), a T2-weighted image (T2WI) and a proton density-weighted image (PDWI) were used for numerical evaluation. Sampling ratio of the Cartesian grid was 30%. Reconstruction was performed by the projection onto convex sets (POCS) method with soft thresholding, subject to data fidelity constraints. In the absence of noise, RMSE of 3D-T1WI was 1.50, ant that of the 2D-T1WI of the transverse plane was in the range of 1.06 to 1.54; anatomical ROIs was in the range of 0.75 to 2.80; those of T2WI were 3.20, 2.77 to 3.06, and 1.81 to 4.51; those of PDWI were 1.69, 1.33 to 1.49, and 1.08 to 1.86. Visual evaluation was performed by three radiologists on the basis of three categories: artifact, anatomical structure, and tissue contrast. Average score of the visual evaluation indicated that anatomical structure and tissue contrast of CS images were equal to those of the original image, although a few artifacts were visible. If noise level was assumed to be 20 dB or less, anatomical structure and tissue contrast were not significantly degraded compared to noise-free CS images.

[Numerical and Visual Evaluations of Compressed Sensing MRI Using 2D Radial Sampling]

Two-dimensional radial MRI using compressed sensing (2D radial CS) enables incoherence sampling in k space unlike conventional Cartesian MRI, however 2D radial CS has not been sufficiently investigated. Numerical and visual evaluations of 2D radial CS were performed in this paper. Three brain anatomical ROIs (white matter, gray matter, cerebrospinal fluid) of a T1-weigthted image (T1WI), a T2-weighted image (T2WI) and a proton density-weighted image (PDWI) were used for the numerical evaluation. The Brainweb MRI Data Base was used for test images. Projection of 80 spokes with linear sampling of 256 pixels was used. Reconstruction was performed by minimizing the L1 norm of a transformed image using wavelet transform and spatial finite-differences (total variation), subject to data fidelity constraint. In the absence of noise, the root mean square error (RMSE) of T1WI was in the range of 3.75 to 5.05; that of the anatomical region of interests (ROIs) was in the range of 1.54 to 10.24; those of T2WI were 8.75 to 11.65 and 4.31 to 6.99; and those of PDWI were 3.44 to 4.46 and 1.34 to 3.09. Visual evaluation was performed by three radiologists on the basis of three categories: artifact, anatomical structure, and tissue contrast. Average percent scores of the visual evaluation were 96% for T1WI, 74-81% for T2WI, and 81-89% for PDWI.

FULLY CONVOLUTIONAL NETWORKS FOR MULTI-MODALITY ISOINTENSE INFANT BRAIN IMAGE SEGMENTATION

The segmentation of infant brain tissue images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) plays an important role in studying early brain development. In the isointense phase (approximately 6-8 months of age), WM and GM exhibit similar levels of intensity in both T1 and T2 MR images, resulting in extremely low tissue contrast and thus making the tissue segmentation very challenging. The existing methods for tissue segmentation in this isointense phase usually employ patch-based sparse labeling on single T1, T2 or fractional anisotropy (FA) modality or their simply-stacked combinations without fully exploring the multi-modality information. To address the challenge, in this paper, we propose to use fully convolutional networks (FCNs) for the segmentation of isointense phase brain MR images. Instead of simply stacking the three modalities, we train one network for each modality image, and then fuse their high-layer features together for final segmentation. Specifically, we conduct a convolution-pooling stream for multimodality information from T1, T2, and FA images separately, and then combine them in high-layer for finally generating the segmentation maps as the outputs. We compared the performance of our approach with that of the commonly used segmentation methods on a set of manually segmented isointense phase brain images. Results showed that our proposed model significantly outperformed previous methods in terms of accuracy. In addition, our results also indicated a better way of integrating multi-modality images, which leads to performance improvement.

Annual Research Review: Discovery science strategies in studies of the pathophysiology of child and adolescent psychiatric disorders--promises and limitations

BACKGROUND

Psychiatric science remains descriptive, with a categorical nosology intended to enhance interobserver reliability. Increased awareness of the mismatch between categorical classifications and the complexity of biological systems drives the search for novel frameworks including discovery science in Big Data. In this review, we provide an overview of incipient approaches, primarily focused on classically categorical diagnoses such as schizophrenia (SZ), autism spectrum disorder (ASD), and attention-deficit/hyperactivity disorder (ADHD), but also reference convincing, if focal, advances in cancer biology, to describe the challenges of Big Data and discovery science, and outline approaches being formulated to overcome existing obstacles.

FINDINGS

A paradigm shift from categorical diagnoses to a domain/structure-based nosology and from linear causal chains to complex causal network models of brain-behavior relationship is ongoing. This (r)evolution involves appreciating the complexity, dimensionality, and heterogeneity of neuropsychiatric data collected from multiple sources ('broad' data) along with data obtained at multiple levels of analysis, ranging from genes to molecules, cells, circuits, and behaviors ('deep' data). Both of these types of Big Data landscapes require the use and development of robust and powerful informatics and statistical approaches. Thus, we describe Big Data analysis pipelines and the promise and potential limitations in using Big Data approaches to study psychiatric disorders.

CONCLUSIONS

We highlight key resources available for psychopathological studies and call for the application and development of Big Data approaches to dissect the causes and mechanisms of neuropsychiatric disorders and identify corresponding biomarkers for early diagnosis.

Manganese-enhanced MRI of rat brain based on slow cerebral delivery of manganese(II) with silica-encapsulated Mn x Fe(1-x) O nanoparticles

In this work, we report a monodisperse bifunctional nanoparticle system, MIO@SiO2 -RITC, as an MRI contrast agent [core, manganese iron oxide (MIO); shell, amorphous silica conjugated with rhodamine B isothiocyanate (RITC)]. It was prepared by thermal decomposition and modified microemulsion methods. The nanoparticles with varying iron to manganese ratios displayed different saturated magnetizations and relaxivities. In vivo MRI of rats injected intravenously with MIO@SiO2-RITC nanoparticles exhibited enhancement of the T1 contrast in brain tissue, in particular a time-delayed enhancement in the hippocampus, pituitary gland, striatum and cerebellum. This is attributable to the gradual degradation of MIO@SiO2-RITC nanoparticles in the liver, resulting in the slow release of manganese(II) [Mn(II)] into the blood pool and, subsequently, accumulation in the brain tissue. Thus, T1-weighted contrast enhancement was clearly detected in the anatomic structure of the brain as time progressed. In addition, T2*-weighted images of the liver showed a gradual darkening effect. Here, we demonstrate the concept of the slow release of Mn(II) for neuroimaging. This new nanoparticle-based manganese contrast agent allows one simple intravenous injection (rather than multiple infusions) of Mn(II) precursor, and results in delineation of the detailed anatomic neuroarchitecture in MRI; hence, this provides the advantage of the long-term study of neural function.

Symmetric inverse consistent nonlinear registration driven by mutual information

A nonlinear viscoelastic image registration algorithm based on the demons paradigm and incorporating inverse consistent constraint (ICC) is implemented. An inverse consistent and symmetric cost function using mutual information (MI) as a similarity measure is employed. The cost function also includes regularization of transformation and inverse consistent error (ICE). The uncertainties in balancing various terms in the cost function are avoided by alternatively minimizing the similarity measure, the regularization of the transformation, and the ICE terms. The diffeomorphism of registration for preventing folding and/or tearing in the deformation is achieved by the composition scheme. The quality of image registration is first demonstrated by constructing brain atlas from 20 adult brains (age range 30-60). It is shown that with this registration technique: (1) the Jacobian determinant is positive for all voxels and (2) the average ICE is around 0.004 voxels with a maximum value below 0.1 voxels. Further, the deformation-based segmentation on Internet Brain Segmentation Repository, a publicly available dataset, has yielded high Dice similarity index (DSI) of 94.7% for the cerebellum and 74.7% for the hippocampus, attesting to the quality of our registration method.