اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدSex differences in predicting fluid intelligence of adolescent brain from T1-weighted MRIs
225
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Fluid intelligence (Gf) has been defined as the ability to reason and solve previously unseen problems. Links to Gf have been found in magnetic resonance imaging (MRI) sequences such as functional MRI and diffusion tensor imaging. As part of the Adolescent Brain Cognitive Development Neurocognitive Prediction Challenge 2019, we sought to predict Gf in children aged 9-10 from T1-weighted (T1W) MRIs. The data included atlas-aligned volumetric T1W images, atlas-defined segmented regions, age, and sex for 3739 subjects used for training and internal validation and 415 subjects used for external validation. We trained sex-specific convolutional neural net (CNN) and random forest models to predict Gf. For the convolutional model, skull-stripped volumetric T1W images aligned to the SRI24 brain atlas were used for training. Volumes of segmented atlas regions along with each subjects age were used to train the random forest regressor models. Performance was measured using the mean squared error (MSE) of the predictions. Random forest models achieved lower MSEs than CNNs. Further, the external validation data had a better MSE for females than males (60.68 vs. 80.74), with a combined MSE of 70.83. Our results suggest that predictive models of Gf from volumetric T1W MRI features alone may perform better when trained separately on male and female data. However, the performance of our models indicates that more information is necessary beyond the available data to make accurate predictions of Gf.
قيم البحث
اقرأ أيضاً
In this work, we utilize T1-weighted MR images and StackNet to predict fluid intelligence in adolescents. Our framework includes feature extraction, feature normalization, feature denoising, feature selection, training a StackNet, and predicting flui
d intelligence. The extracted feature is the distribution of different brain tissues in different brain parcellation regions. The proposed StackNet consists of three layers and 11 models. Each layer uses the predictions from all previous layers including the input layer. The proposed StackNet is tested on a public benchmark Adolescent Brain Cognitive Development Neurocognitive Prediction Challenge 2019 and achieves a mean squared error of 82.42 on the combined training and validation set with 10-fold cross-validation. In addition, the proposed StackNet also achieves a mean squared error of 94.25 on the testing data. The source code is available on GitHub.
The relation between spontaneous and stimulated global brain activity is a fundamental problem in the understanding of brain functions. This question is investigated both theoretically and experimentally within the context of nonequilibrium fluctuati
on-dissipation relations. We consider the stochastic coarse-grained Wilson-Cowan model in the linear noise approximation and compare analytical results to experimental data from magnetoencephalography (MEG) of human brain. The short time behavior of the autocorrelation function for spontaneous activity is characterized by a double-exponential decay, with two characteristic times, differing by two orders of magnitude. Conversely, the response function exhibits a single exponential decay in agreement with experimental data for evoked activity under visual stimulation. Results suggest that the brain response to weak external stimuli can be predicted from the observation of spontaneous activity and pave the way to controlled experiments on the brain response under different external perturbations.
Brain connectivity with functional Magnetic Resonance Imaging (fMRI) is a popular approach for detecting differences between healthy and clinical populations. Before creating a functional brain network, the fMRI time series must undergo several prepr
ocessing steps to control for artifacts and to improve data quality. However, preprocessing may affect the results in an undesirable way. Spatial smoothing, for example, is known to alter functional network structure. Yet, its effects on group-level network differences remain unknown. Here, we investigate the effects of spatial smoothing on the difference between patients and controls for two clinical conditions: autism spectrum disorder and bipolar disorder, considering fMRI data smoothed with Gaussian kernels (0-32 mm). We find that smoothing affects network differences between groups. For weighted networks, incrementing the smoothing kernel makes networks more different. For thresholded networks, larger smoothing kernels lead to more similar networks, although this depends on the network density. Smoothing also alters the effect sizes of the individual link differences. This is independent of the ROI size, but vary with link length. The effects of spatial smoothing are diverse, non-trivial, and difficult to predict. This has important consequences: the choice of smoothing kernel affects the observed network differences.
The deep neural nets of modern artificial intelligence (AI) have not achieved defining features of biological intelligence, including abstraction, causal learning, and energy-efficiency. While scaling to larger models has delivered performance improv
ements for current applications, more brain-like capacities may demand new theories, models, and methods for designing artificial learning systems. Here, we argue that this opportunity to reassess insights from the brain should stimulate cooperation between AI research and theory-driven computational neuroscience (CN). To motivate a brain basis of neural computation, we present a dynamical view of intelligence from which we elaborate concepts of sparsity in network structure, temporal dynamics, and interactive learning. In particular, we suggest that temporal dynamics, as expressed through neural synchrony, nested oscillations, and flexible sequences, provide a rich computational layer for reading and updating hierarchical models distributed in long-term memory networks. Moreover, embracing agent-centered paradigms in AI and CN will accelerate our understanding of the complex dynamics and behaviors that build useful world models. A convergence of AI/CN theories and objectives will reveal dynamical principles of intelligence for brains and engineered learning systems. This article was inspired by our symposium on dynamical neuroscience and machine learning at the 6th Annual US/NIH BRAIN Initiative Investigators Meeting.
The dynamic characteristics of functional network connectivity have been widely acknowledged and studied. Both shared and unique information has been shown to be present in the connectomes. However, very little has been known about whether and how th
is common pattern can predict the individual variability of the brain, i.e. brain fingerprinting, which attempts to reliably identify a particular individual from a pool of subjects. In this paper, we propose to enhance the individual uniqueness based on an autoencoder network. More specifically, we rely on the hypothesis that the common neural activities shared across individuals may lessen individual discrimination. By reducing contributions from shared activities, inter-subject variability can be enhanced. Results show that that refined connectomes utilizing an autoencoder with sparse dictionary learning can successfully distinguish one individual from the remaining participants with reasonably high accuracy (up to 99:5% for the rest-rest pair). Furthermore, high-level cognitive behavior (e.g., fluid intelligence, executive function, and language comprehension) can also be better predicted using refined functional connectivity profiles. As expected, the high-order association cortices contributed more to both individual discrimination and behavior prediction. The proposed approach provides a promising way to enhance and leverage the individualized characteristics of brain networks.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
