End-to-End Multi-task Learning Architecture for Brain Tumor Analysis with Uncertainty Estimation in MRI Images.

Brain tumors are a threat to life for every other human being, be it adults or children. Gliomas are one of the deadliest brain tumors with an extremely difficult diagnosis. The reason is their complex and heterogenous structure which gives rise to subjective as well as objective errors. Their manual segmentation is a laborious task due to their complex structure and irregular appearance. To cater to all these issues, a lot of research has been done and is going on to develop AI-based solutions that can help doctors and radiologists in the effective diagnosis of gliomas with the least subjective and objective errors, but an end-to-end system is still missing. An all-in-one framework has been proposed in this research. The developed end-to-end multi-task learning (MTL) architecture with a feature attention module can classify, segment, and predict the overall survival of gliomas by leveraging task relationships between similar tasks. Uncertainty estimation has also been incorporated into the framework to enhance the confidence level of healthcare practitioners. Extensive experimentation was performed by using combinations of MRI sequences. Brain tumor segmentation (BraTS) challenge datasets of 2019 and 2020 were used for experimental purposes. Results of the best model with four sequences show 95.1% accuracy for classification, 86.3% dice score for segmentation, and a mean absolute error (MAE) of 456.59 for survival prediction on the test data. It is evident from the results that deep learning-based MTL models have the potential to automate the whole brain tumor analysis process and give efficient results with least inference time without human intervention. Uncertainty quantification confirms the idea that more data can improve the generalization ability and in turn can produce more accurate results with less uncertainty. The proposed model has the potential to be utilized in a clinical setup for the initial screening of glioma patients.

Robust and adaptive terrain classification and gait event detection system.

Real-time gait event detection (GED) system can be utilized for gait analysis and tracking fitness activities. GED for various types of terrains (e.g., stair-walk, uneven surfaces, etc.) is still an open research problem. This study presents an inertial sensor-based approach for real-time GED system that works for diverse terrains in an uncontrolled environment. The GED system classifies three types of terrains, i.e., flat-walk, stair-ascend and stair-descend, with an average classification accuracy of 99%. It also accurately detects various gait events, including, toe-strike, heel-rise, toe-off, and heel-strike. It is computationally efficient, implemented on a low-cost microcontroller, works in real-time and can be used in portable rehabilitation devices for use in dynamic environments.

Sentiments analysis of fMRI using automatically generated stimuli labels under naturalistic paradigm.

Our emotions and sentiments are influenced by naturalistic stimuli such as the movies we watch and the songs we listen to, accompanied by changes in our brain activation. Comprehension of these brain-activation dynamics can assist in identification of any associated neurological condition such as stress and depression, leading towards making informed decision about suitable stimuli. A large number of open-access functional magnetic resonance imaging (fMRI) datasets collected under naturalistic conditions can be used for classification/prediction studies. However, these datasets do not provide emotion/sentiment labels, which limits their use in supervised learning studies. Manual labeling by subjects can generate these labels, however, this method is subjective and biased. In this study, we are proposing another approach of generating automatic labels from the naturalistic stimulus itself. We are using sentiment analyzers (VADER, TextBlob, and Flair) from natural language processing to generate labels using movie subtitles. Subtitles generated labels are used as the class labels for positive, negative, and neutral sentiments for classification of brain fMRI images. Support vector machine, random forest, decision tree, and deep neural network classifiers are used. We are getting reasonably good classification accuracy (42-84%) for imbalanced data, which is increased (55-99%) for balanced data.

Role of deep learning in brain tumor detection and classification (2015 to 2020): A review.

During the last decade, computer vision and machine learning have revolutionized the world in every way possible. Deep Learning is a sub field of machine learning that has shown remarkable results in every field especially biomedical field due to its ability of handling huge amount of data. Its potential and ability have also been applied and tested in the detection of brain tumor using MRI images for effective prognosis and has shown remarkable performance. The main objective of this research work is to present a detailed critical analysis of the research and findings already done to detect and classify brain tumor through MRI images in the recent past. This analysis is specifically beneficial for the researchers who are experts of deep learning and are interested to apply their expertise for brain tumor detection and classification. As a first step, a brief review of the past research papers using Deep Learning for brain tumor classification and detection is carried out. Afterwards, a critical analysis of Deep Learning techniques proposed in these research papers (2015-2020) is being carried out in the form of a Table. Finally, the conclusion highlights the merits and demerits of deep neural networks. The results formulated in this paper will provide a thorough comparison of recent studies to the future researchers, along with the idea of the effectiveness of various deep learning approaches. We are confident that this study would greatly assist in advancement of brain tumor research.

Classification of forearm EMG signals for 10 motions using optimum feature-channel combinations.

Electromyography (EMG) is the study of electrical activity in the muscles. We classify EMG signals from surface electrodes (channels) using Artificial Neural Network (ANN). We evaluate classification performance of 10 different hand motions using several feature-channel combinations with wrapper method. Highest classification accuracy of 98.7% is achieved with each feature-channel combination. Compared to previous studies, we achieve the highest accuracy for 10 classes with lower number of feature-channel combination. We reduce ANN complexity without compromising the classification accuracy for deployment in low-end hardware with limited computational power along with improving the design of a low-cost hardware for EMG signal acquisition.

Brain Injury and Dementia in Pakistan: Current Perspectives.

Alzheimer's disease (AD) is the most common form of dementia, accounting for 50-75% of all cases, with a greater proportion of individuals affected at older age range. A single moderate or severe traumatic brain injury (TBI) is associated with accelerated aging and increased risk for dementia. The fastest growth in the elderly population is taking place in China, Pakistan, and their south Asian neighbors. Current clinical assessments are based on data collected from Caucasian populations from wealthy backgrounds giving rise to a "diversity" crisis in brain research. Pakistan is a lower-middle income country (LMIC) with an estimated one million people living with dementia. Pakistan also has an amalgamation of risk factors that lead to brain injuries such as lack of road legislations, terrorism, political instability, and domestic and sexual violence. Here, we provide an initial and current assessment of the incidence and management of dementia and TBI in Pakistan. Our review demonstrates the lack of resources in terms of speciality trained clinician staff, medical equipment, research capabilities, educational endeavors, and general awareness in the fields of dementia and TBI. Pakistan also lacks state-of-the-art assessment of dementia and its risk factors, such as neuroimaging of brain injury and aging. We provide recommendations for improvement in this arena that include the recent creation of Pakistan Brain Injury Consortium (PBIC). This consortium will enhance international collaborative efforts leading to capacity building for innovative research, clinician and research training and developing databases to bring Pakistan into the international platform for dementia and TBI research.

Parametric Dependencies of Sliding Window Correlation.

OBJECTIVE: In this paper, we explore the dependence of sliding window correlation (SWC) results on different parameters of correlating signals. The SWC is extensively used to explore the dynamics of functional connectivity (FC) networks using resting-state functional MRI (rsfMRI) scans. These scanned signals often contain multiple amplitudes, frequencies, and phases. However, the exact values of these parameters are unknown. Two recent studies explored the relationship of window length and frequencies (minimum/maximum) in the correlating signals. METHODS: We extend the findings of these studies by using two deterministic signals with multiple amplitudes, frequencies, and phases. Afterward, we modulate one of the signals to introduce dynamics (nonstationarity) in their relationship. We also explore the relationship of window length and frequency band for real rsfMRI data. RESULTS: For deterministic signals, the spurious fluctuations due to the method itself minimize, and the SWC estimates the stationary correlation when frequencies in the signals have specific relationship. For dynamic relationship also, the undesirable frequencies were removed under specific conditions for the frequencies. For real rsfMRI data, the SWC results varied with frequencies and window length. CONCLUSION: In the absence of any "ground truth" for different parameters in real rsfMRI signals, the SWC with a constant window size may not be a reliable method to study the dynamics of the FC. SIGNIFICANCE: This study reveals the parametric dependencies of the SWC and its limitation as a method to analyze dynamics of FC networks in the absence of any ground truth.

Instantaneous brain dynamics mapped to a continuous state space.

Measures of whole-brain activity, from techniques such as functional Magnetic Resonance Imaging, provide a means to observe the brain's dynamical operations. However, interpretation of whole-brain dynamics has been stymied by the inherently high-dimensional structure of brain activity. The present research addresses this challenge through a series of scale transformations in the spectral, spatial, and relational domains. Instantaneous multispectral dynamics are first developed from input data via a wavelet filter bank. Voxel-level signals are then projected onto a representative set of spatially independent components. The correlation distance over the instantaneous wavelet-ICA state vectors is a graph that may be embedded onto a lower-dimensional space to assist the interpretation of state-space dynamics. Applying this procedure to a large sample of resting-state and task-active data (acquired through the Human Connectome Project), we segment the empirical state space into a continuum of stimulus-dependent brain states. Upon observing the local neighborhood of brain-states adopted subsequent to each stimulus, we may conclude that resting brain activity includes brain states that are, at times, similar to those adopted during tasks, but that are at other times distinct from task-active brain states. As task-active brain states often populate a local neighborhood, back-projection of segments of the dynamical state space onto the brain's surface reveals the patterns of brain activity that support many experimentally-defined states.

Noise and non-neuronal contributions to the BOLD signal: applications to and insights from animal studies.

The BOLD signal reflects hemodynamic events within the brain, which in turn are driven by metabolic changes and neural activity. However, the link between BOLD changes and neural activity is indirect and can be influenced by a number of non-neuronal processes. Motion and physiological cycles have long been known to affect the BOLD signal and are present in both humans and animal models. Differences in physiological baseline can also contribute to intra- and inter-subject variability. The use of anesthesia, common in animal studies, alters neural activity, vascular tone, and neurovascular coupling. Most intriguing, perhaps, are the contributions from other processes that do not appear to be neural in origin but which may provide information about other aspects of neurophysiology. This review discusses different types of noise and non-neuronal contributors to the BOLD signal, sources of variability for animal studies, and insights to be gained from animal models.

Evaluation of sliding window correlation performance for characterizing dynamic functional connectivity and brain states.

A promising recent development in the study of brain function is the dynamic analysis of resting-state functional MRI scans, which can enhance understanding of normal cognition and alterations that result from brain disorders. One widely used method of capturing the dynamics of functional connectivity is sliding window correlation (SWC). However, in the absence of a "gold standard" for comparison, evaluating the performance of the SWC in typical resting-state data is challenging. This study uses simulated networks (SNs) with known transitions to examine the effects of parameters such as window length, window offset, window type, noise, filtering, and sampling rate on the SWC performance. The SWC time course was calculated for all node pairs of each SN and then clustered using the k-means algorithm to determine how resulting brain states match known configurations and transitions in the SNs. The outcomes show that the detection of state transitions and durations in the SWC is most strongly influenced by the window length and offset, followed by noise and filtering parameters. The effect of the image sampling rate was relatively insignificant. Tapered windows provide less sensitivity to state transitions than rectangular windows, which could be the result of the sharp transitions in the SNs. Overall, the SWC gave poor estimates of correlation for each brain state. Clustering based on the SWC time course did not reliably reflect the underlying state transitions unless the window length was comparable to the state duration, highlighting the need for new adaptive window analysis techniques.

Multiscale network activity in resting state fMRI.

The brain is inherently multiscalar in both space and time. We argue that this multiscalar nature is reflected in the blood oxygenation level dependent (BOLD) fluctuations used to map functional connectivity. We present evidence that global fluctuations in activity, quasiperiodic spatiotemporal patterns, and aperiodic time-varying activity coexist within the BOLD signal. These processes can be separated using careful analysis and appear to reflect electrical activity on similar scales, suggesting that the BOLD signal fluctuations can provide novel insight into the functional architecture of the brain.

Adaptive change point detection of dynamic functional connectivity networks.

This study presents a new algorithm to adaptively detect change points of functional connectivity networks in the brain. It uses scans from resting-state functional magnetic resonance imaging (rsfMRI) which is one of the major tools to investigate intrinsic brain functionality. Different regions of the resting brain form networks that change states within a few seconds to minutes. The change points of these networks are different in normal and disordered brain functions and their understanding can help in identification of brain disorders. These changes arise from many unknown factors and extraction of these change points is one of the the major challenges in the absence of any ground truth. Our algorithm detects these change points adaptively by computing sum of absolute sign differences of adjacent images in rsfMRI scans using measures from image and video processing. We demonstrate the effectiveness of the proposed algorithm and show that these change points can be detected reliably in both task-based and resting-state networks. The outcomes also point to new directions for future work.

Considerations for resting state functional MRI and functional connectivity studies in rodents.

Resting state functional MRI (rs-fMRI) and functional connectivity mapping have become widely used tools in the human neuroimaging community and their use is rapidly spreading into the realm of rodent research as well. One of the many attractive features of rs-fMRI is that it is readily translatable from humans to animals and back again. Changes in functional connectivity observed in human studies can be followed by more invasive animal experiments to determine the neurophysiological basis for the alterations, while exploratory work in animal models can identify possible biomarkers for further investigation in human studies. These types of interwoven human and animal experiments have a potentially large impact on neuroscience and clinical practice. However, impediments exist to the optimal application of rs-fMRI in small animals, some similar to those encountered in humans and some quite different. In this review we identify the most prominent of these barriers, discuss differences between rs-fMRI in rodents and in humans, highlight best practices for animal studies, and review selected applications of rs-fMRI in rodents. Our goal is to facilitate the integration of human and animal work to the benefit of both fields.

Phase-amplitude coupling and infraslow (<1 Hz) frequencies in the rat brain: relationship to resting state fMRI.

Resting state functional magnetic resonance imaging (fMRI) can identify network alterations that occur in complex psychiatric diseases and behaviors, but its interpretation is difficult because the neural basis of the infraslow BOLD fluctuations is poorly understood. Previous results link dynamic activity during the resting state to both infraslow frequencies in local field potentials (LFP) (<1 Hz) and band-limited power in higher frequency LFP (>1 Hz). To investigate the relationship between these frequencies, LFPs were recorded from rats under two anesthetics: isoflurane and dexmedetomidine. Signal phases were calculated from low-frequency LFP and compared to signal amplitudes from high-frequency LFP to determine if modulation existed between the two frequency bands (phase-amplitude coupling). Isoflurane showed significant, consistent phase-amplitude coupling at nearly all pairs of frequencies, likely due to the burst-suppression pattern of activity that it induces. However, no consistent phase-amplitude coupling was observed in rats that were anesthetized with dexmedetomidine. fMRI-LFP correlations under isoflurane using high frequency LFP were reduced when the low frequency LFP's influence was accounted for, but not vice-versa, or in any condition under dexmedetomidine. The lack of consistent phase-amplitude coupling under dexmedetomidine and lack of shared variance between high frequency and low frequency LFP as it relates to fMRI suggests that high and low frequency neural electrical signals may contribute differently, possibly even independently, to resting state fMRI. This finding suggests that researchers take care in interpreting the neural basis of resting state fMRI, as multiple dynamic factors in the underlying electrophysiology could be driving any particular observation.

Cluster-based analysis for characterizing dynamic functional connectivity.

Different regions in the resting brain exhibit non-stationary functional connectivity (FC) over time. In this paper, a simple and efficient framework of clustering the variability in FC of a rat's brain at rest is proposed. This clustering process reveals areas that are always connected with a chosen region, called seed voxel, along with the areas exhibiting variability in the FC. This addresses an issue common to most dynamic FC analysis techniques, which is the assumption that the spatial extent of a given network remains constant over time. We increase the voxel size and reduce the spatial resolution to analyze variable FC of the whole resting brain. We hypothesize that the adjacent voxels in resting state functional magnetic resonance imaging (rsfMRI), just as in task-based fMRI, exhibit similar intensities, so they can be averaged to obtain larger voxels without any significant loss of information. Sliding window correlation is used to compute variable patterns of the rat's whole brain FC with the seed voxel in the sensorimotor cortex. These patterns are grouped based on their spatial similarities using binary transformed feature vectors in k-means clustering, not only revealing the variable and nonvariable portions of FC in the resting brain but also detecting the extent of the variability of these patterns.

Monoclonal behavior of molecularly imprinted polymer nanoparticles in capillary electrochromatography.

A new approach based on miniemulsion polymerization is demonstrated for synthesis of molecularly imprinted nanoparticles (MIP-NP; 30-150 nm) with "monoclonal" binding behavior. The performance of the MIP nanoparticles is characterized with partial filling capillary electrochromatography, for the analysis of rac-propranolol, where (S)-propranolol is used as a template. In contrast to previous HPLC and CEC methods based on the use of MIPs, there is no apparent tailing for the enantiomer peaks, and baseline separation with 25,000-60,000 plate number is achieved. These effects are attributed to reduction of the MIP site heterogeneity by means of peripheral location of the core cross-linked NP and to MIP-binding sites with the same ordered radial orientation. This new MIP approach is based on the substitution of the functional monomers with a surfactant monomer, sodium N-undecenoyl glycinate (SUG) for improved inclusion in the MIP-NP structure and to the use of a miniemulsion in the MIP-NP synthesis. The feasibility of working primarily with aqueous electrolytes (10 mM phosphate with a 20% acetonitrile at pH 7) is attributable to the micellar character of the MIP-NPs, provided by the inclusion of the SUG monomers in the structure. To our knowledge this is the first example of "monoclonal" MIP-NPs incorporated in CEC separations of drug enantiomers.