fMRI-Targeted High-Angular Resolution Diffusion MR Tractography to Identify Functional Language Tracts in Healthy Controls and Glioma Patients

Francesco Sanvito Edit "fMRI-Targeted High-Angular Resolution Diffusion MR Tractography to Identify Functional Language Tracts in Healthy Controls and Glioma Patients"
Background

MR Tractography enables non-invasive preoperative depiction of language subcortical tracts, which is crucial for the presurgical work-up of brain tumors; however, it cannot evaluate the exact function of the fibers.

Purpose

A systematic pipeline was developed to combine tractography reconstruction of language fiber bundles, based on anatomical landmarks (Anatomical-T), with language fMRI cortical activations. A fMRI-targeted Tractography (fMRI-T) was thus obtained, depicting the subsets of the anatomical tracts whose endpoints are located inside a fMRI activation. We hypothesized that fMRI-T could provide additional functional information regarding the subcortical structures, better reflecting the eloquent white matter structures identified intraoperatively.

Methods

Both Anatomical-T and fMRI-T of language fiber tracts were performed on 16 controls and preoperatively on 16 patients with left-hemisphere brain tumors, using a q-ball residual bootstrap algorithm based on High Angular Resolution Diffusion Imaging (HARDI) datasets (b = 3000 s/mm2; 60 directions); fMRI ROIs were obtained using picture naming, verbal fluency, and auditory verb generation tasks. In healthy controls, normalized MNI atlases of fMRI-T and Anatomical-T were obtained. In patients, the surgical resection of the tumor was pursued by identifying eloquent structures with intraoperative direct electrical stimulation mapping and extending surgery to the functional boundaries. Post-surgical MRI allowed to identify Anatomical-T and fMRI-T non-eloquent portions removed during the procedure.

Results

MNI Atlases showed that fMRI-T is a subset of Anatomical-T, and that different task-specific fMRI-T involve both shared subsets and task-specific subsets – e.g., verbal fluency fMRI-T strongly involves dorsal frontal tracts, consistently with the phonogical-articulatory features of this task. A quantitative analysis in patients revealed that Anatomical-T removed portions of AF-SLF and IFOF were significantly greater than verbal fluency fMRI-T ones, suggesting that fMRI-T is a more specific approach. In addition, qualitative analyses showed that fMRI-T AF-SLF and IFOF predict the exact functional limits of resection with increased specificity when compared to Anatomical-T counterparts, especially the superior frontal portion of IFOF, in a subcohort of patients.

Conclusion

These results suggest that performing fMRI-T in addition to the ‘classic’ Anatomical-T may be useful in a preoperative setting to identify the ‘high-risk subsets’ that should be spared during the surgical procedure.

Higher Sensitivity and Reproducibility of Wavelet-Based Amplitude of Resting-State fMRI

Fei-Fei Luo Edit "Higher Sensitivity and Reproducibility of Wavelet-Based Amplitude of Resting-State fMRI"

The fast Fourier transform (FFT) is a widely used algorithm used to depict the amplitude of low-frequency fluctuation (ALFF) of resting-state functional magnetic resonance imaging (RS-fMRI). Wavelet transform (WT) is more effective in representing the complex waveform due to its adaptivity to non-stationary or local features of data and many varieties of wavelet functions with different shapes being available. However, there is a paucity of RS-fMRI studies that systematically compare between the results of FFT versus WT. The present study employed five cohorts of datasets and compared the sensitivity and reproducibility of FFT-ALFF with those of Wavelet-ALFF based on five mother wavelets (namely, db2, bior4.4, morl, meyr, and sym3). In addition to the conventional frequency band of 0.0117–0.0781 Hz, a comparison was performed in sub-bands, namely, Slow-6 (0–0.0117 Hz), Slow-5 (0.0117–0.0273 Hz), Slow-4 (0.0273–0.0742 Hz), Slow-3 (0.0742–0.1992 Hz), and Slow-2 (0.1992–0.25 Hz). The results indicated that the Wavelet-ALFF of all five mother wavelets was generally more sensitive and reproducible than FFT-ALFF in all frequency bands. Specifically, in the higher frequency band Slow-2 (0.1992–0.25 Hz), the mean sensitivity of db2-ALFF results was 1.54 times that of FFT-ALFF, and the reproducibility of db2-ALFF results was 2.95 times that of FFT-ALFF. The findings suggest that wavelet-ALFF can replace FFT-ALFF, especially in the higher frequency band. Future studies should test more mother wavelets on other RS-fMRI metrics and multiple datasets.

Neurodegenerative Diseases – Is Metabolic Deficiency the Root Cause?

Vignayanandam Ravindernath Muddapu Edit "Neurodegenerative Diseases – Is Metabolic Deficiency the Root Cause?"

Neurodegenerative diseases, including Alzheimer, Parkinson, Huntington, and amyotrophic lateral sclerosis, are a prominent class of neurological diseases currently without a cure. They are characterized by an inexorable loss of a specific type of neurons. The selective vulnerability of specific neuronal clusters (typically a subcortical cluster) in the early stages, followed by the spread of the disease to higher cortical areas, is a typical pattern of disease progression. Neurodegenerative diseases share a range of molecular and cellular pathologies, including protein aggregation, mitochondrial dysfunction, glutamate toxicity, calcium load, proteolytic stress, oxidative stress, neuroinflammation, and aging, which contribute to neuronal death. Efforts to treat these diseases are often limited by the fact that they tend to address any one of the above pathological changes while ignoring others. Lack of clarity regarding a possible root cause that underlies all the above pathologies poses a significant challenge. In search of an integrative theory for neurodegenerative pathology, we hypothesize that metabolic deficiency in certain vulnerable neuronal clusters is the common underlying thread that links many dimensions of the disease. The current review aims to present an outline of such an integrative theory. We present a new perspective of neurodegenerative diseases as metabolic disorders at molecular, cellular, and systems levels. This helps to understand a common underlying mechanism of the many facets of the disease and may lead to more promising disease-modifying therapeutic interventions. Here, we briefly discuss the selective metabolic vulnerability of specific neuronal clusters and also the involvement of glia and vascular dysfunctions. Any failure in satisfaction of the metabolic demand by the neurons triggers a chain of events that precipitate various manifestations of neurodegenerative pathology.

Alpha down-regulation neurofeedback training effects on implicit motor learning and consolidation

Wenya Nan, Limin Yang, Feng Wan, Frank Zhu and Yong Hu Edit "Alpha down-regulation neurofeedback training effects on implicit motor learning and consolidation"
Objective. Implicit motor learning, which is a non-conscious form of learning characterized by motor performance improvement with practice, plays an essential role in various daily activities. Earlier study using neurofeedback training (NFT), a type of brain-computer interaction that enables the user to learn self-regulating his/her own brain activity, demonstrated that down-regulating alpha over primary motor cortex by NFT could immediately facilitate the implicit motor learning in a relatively simple motor task. However, detailed effects on EEG and implicit motor learning due to NFT especially in a more complex motor task are still unclear. Approach. We designed a single-blind sham-controlled between-subject study to examine whether alpha down-regulation NFT could facilitate implicit motor learning and also its consolidation in a more difficult and motor predominant task. At left primary motor cortex (C3) in two days, the alpha NFT group received alpha down-regu...

Planning with Brain-inspired AI. (arXiv:2003.12353v1 [cs.AI])

Naoya Arakawa Edit "Planning with Brain-inspired AI. (arXiv:2003.12353v1 [cs.AI])"

This article surveys engineering and neuroscientific models of planning as a cognitive function, which is regarded as a typical function of fluid intelligence in the discussion of general intelligence. It aims to present existing planning models as references for realizing the planning function in brain-inspired AI or artificial general intelligence (AGI). It also proposes themes for the research and development of brain-inspired AI from the viewpoint of tasks and architecture.

Going in circles is the way forward: the role of recurrence in visual inference. (arXiv:2003.12128v1 [q-bio.NC])

Ruben S. van Bergen, Nikolaus Kriegeskorte Edit "Going in circles is the way forward: the role of recurrence in visual inference. (arXiv:2003.12128v1 [q-bio.NC])"

Biological visual systems exhibit abundant recurrent connectivity. State-of-the-art neural network models for visual recognition, by contrast, rely heavily or exclusively on feedforward computation. Any finite-time recurrent neural network (RNN) can be unrolled along time to yield an equivalent feedforward neural network (FNN). This important insight suggests that computational neuroscientists may not need to engage recurrent computation, and that computer-vision engineers may be limiting themselves to a special case of FNN if they build recurrent models. Here we argue, to the contrary, that FNNs are a special case of RNNs and that computational neuroscientists and engineers should engage recurrence to understand how brains and machines can (1) achieve greater and more flexible computational depth, (2) compress complex computations into limited hardware, (3) integrate priors and priorities into visual inference through expectation and attention, (4) exploit sequential dependencies in their data for better inference and prediction, and (5) leverage the power of iterative computation.

Bayesian network classifiers using ensembles and smoothing

Latest Results for Knowledge and Information Systems Edit "Bayesian network classifiers using ensembles and smoothing"

Abstract

Bayesian network classifiers are, functionally, an interesting class of models, because they can be learnt out-of-core, i.e. without needing to hold the whole training data in main memory. The selective K-dependence Bayesian network classifier (SKDB) is state of the art in this class of models and has shown to rival random forest (RF) on problems with categorical data. In this paper, we introduce an ensembling technique for SKDB, called ensemble of SKDB (ESKDB). We show that ESKDB significantly outperforms RF on categorical and numerical data, as well as rivalling XGBoost. ESKDB combines three main components: (1) an effective strategy to vary the networks that is built by single classifiers (to make it an ensemble), (2) a stochastic discretization method which allows to both tackle numerical data as well as further increases the variance between different components of our ensemble and (3) a superior smoothing technique to ensure proper calibration of ESKDB’s probabilities. We conduct a large set of experiments with 72 datasets to study the properties of ESKDB (through a sensitivity analysis) and show its competitiveness with the state of the art.